Inhaler system

Abstract

To provide a system for calculating a probability of CORD exacerbation in a subject.SOLUTION: A system comprises a first inhaler for delivering a rescue medicament to a subject. The first inhaler has a use-detection system configured to measure the number of times of rescue inhalation performed by the subject using the first inhaler. The system selectively has a second inhaler for delivering a maintenance medicament to the subject during a routine inhalation. A sensor system is configured to measure a parameter relating to airflow during the rescue inhalation and / or during the routine inhalation when the second inhaler is included in the system. The system further includes a processor configured to measure the number of the rescue inhalations during a first time period and receive the parameter measured for at least some of the rescue inhalations and / or routine inhalations. Using a weighted model, the processor calculates a probability of COPD exacerbation on the basis of the number of rescue inhalations and the parameters.SELECTED DRAWING: Figure 1

Description

Technical Field

[0001] The present invention relates to an inhaler system, and more particularly to a system and method for calculating the probability of exacerbation of a respiratory disease.

Background Art

[0002] Many respiratory diseases, such as asthma or chronic obstructive pulmonary disease (COPD), are lifelong diseases where treatment involves managing the patient's symptoms and lifelong administration of medications to reduce the risk of irreversible lesions. Currently, there is no cure for diseases such as asthma and COPD. Treatment takes two forms. First, in terms of maintenance treatment, it is intended to reduce airway inflammation and as a result, control symptoms in the future. Maintenance treatment is typically carried out by prescribing inhaled corticosteroids alone or in combination with long-acting bronchodilators and / or muscarinic antagonists. Second, treatment also has an emergency (or rescue) treatment aspect, where the patient is administered an immediate-acting bronchodilator to relieve severe attacks of asthma, cough, chest tightness, and dyspnea. Patients suffering from a respiratory disease such as asthma or COPD may also experience a temporary relapse or exacerbation that worsens the symptoms of the respiratory disease. In the worst case, the exacerbation can be life-threatening to the patient.

[0003] If an impending exacerbation of a respiratory disease can be identified, it becomes possible to improve the action plan and provide opportunities for proactive treatment such as unscheduled visits to or from a doctor, hospitalization, and systemic administration of steroids before the patient's symptoms require them.

[0004] Therefore, in the prior art, there is a need for an improved method for identifying the risk of an impending exacerbation of a respiratory disease.

Summary of the Invention

[0005] Accordingly, the present invention provides a system for calculating the probability of exacerbation of COPD in a subject, comprising a first inhaler for administering an emergency drug to the subject, the first inhaler having a usage detection system configured to measure an emergency inhalation performed by the subject using the first inhaler, a selective second inhaler for administering a maintenance drug to the subject during regular inhalation, a sensor system configured to measure parameters related to the airflow during the emergency inhalation and / or during the regular inhalation when the second inhaler is used in the system, a processor configured to measure the number of emergency inhalations within a first period, receive measurement values of the parameters during at least several emergency inhalations and / or during the regular inhalation, and calculate the probability of exacerbation of COPD based on the measurement values of the parameters during the emergency inhalations using a weighted model, wherein the model is weighted such that the measurement values of the parameters are more significant than the number of emergency inhalations in the calculation of the probability.

[0006] By using the number of emergency inhalations and the measurement values of the parameters related to the airflow during emergency and / or regular inhalations, a model for predicting the exacerbation of COPD can be provided, which can predict with higher accuracy than, for example, a model that ignores any of these factors. Furthermore, when predicting the exacerbation of COPD, the parameters related to the airflow during inhalation are made more significant than the number of emergency inhalations in the calculation of the probability. As a result, since the model is weighted such that the inhalation parameters are more significant than the number of emergency inhalations in the calculation of the probability, the accuracy of the calculation of the probability is improved. This is in contrast to the trend for predicting the exacerbation of asthma, in which the number of emergency inhalations is considered more significant than the inhalation parameters in the prediction of exacerbation.

Brief Description of the Drawings

[0007]

Figure 1

Figure 2

Figure 3A

Figure 3B

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Figure 10

Figure 11

Figure 12

Figure 13

Figure 14

Figure 15

Figure 16

Figure 17

Figure 18

Figure 19

Figure 20

Figure 21

Figure 22

Best Mode for Carrying Out the Invention

[0008] It should be understood that the detailed descriptions and specific examples given while exemplifying the device, system and method are for the purpose of merely explaining the present invention and are not intended to limit the scope of the present invention. Those and other features, aspects and advantages of the device, system and method of the present invention will be better understood from the following description of the specification, the appended claims and the drawings. It should be understood that the drawings are only schematic and are not drawn to actual dimensions. Also, it should be understood that the same or similar components are given the same numbers throughout the drawings.

[0009] Asthma and COPD are chronic inflammatory diseases of the airway. Both are characterized by airway obstruction and bronchial paralysis that are prone to change and recur. The symptoms include attacks of wheezing, coughing, chest tightness and dyspnea.

[0010] The symptoms are controlled by avoiding triggers and using medications, particularly inhaled medications. The medications include inhaled corticosteroids and inhaled bronchodilators.

[0011] Inhaled corticosteroids (ICS) are steroid hormone agents used for long-term control of respiratory diseases. They function by reducing airway inflammation. Inhaled corticosteroids include, for example, budesonide, beclomethasone (beclomethasone dipropionate), fluticasone (fluticasone propionate), mometasone (mometasone furoate), ciclesonide and dexamethasone (dexamethasone sodium). The parentheses indicate the preferred salt or ester form.

[0012] Bronchodilators of different classifications target different receptors in the airway. Two commonly used classifications are β 2 agonists and anticholinergics.

[0013] β 2 -adrenergic agonists (β 2 -agonists) act on adrenergic receptors that induce smooth muscle relaxation and dilate the airway. Long-acting β 2 -agonists (LABA) include, for example, formoterol (formoterol fumarate), salmeterol (salmeterol xinafoate), indacaterol (indacaterol maleate), bambuterol (bambuterol hydrochloride), clenbuterol (clenbuterol hydrochloride), olodaterol (olodaterol hydrochloride), carmoterol (carmoterol hydrochloride), tulobuterol (tulobuterol hydrochloride) and vilanterol (vilanterol trifenatate). Short-acting β 2 -agonists (SABA) include, for example, albuterol (albuterol sulfate).

[0014] Typically, short-acting bronchodilators provide rapid relief from acute bronchoconstriction (often referred to as "rescue drugs" or "reliever drugs"), while long-acting bronchodilators assist in long-term symptom control and prevention. However, some long-acting bronchodilators with rapid onset of action, such as formoterol (formoterol fumarate), can be used as rescue medications. That is, rescue medications provide relief from acute bronchoconstriction. Rescue medications are taken as needed. Rescue medications can also be in the form of combination products, such as, for example, ICS-formoterol (formoterol fumarate), typically budesonide-formoterol fumarate dihydrate inhalant. Thus, rescue medications preferably consist of an SABA or a LABA with rapid onset of action, more preferably albuterol (albuterol sulfate) or formoterol (formoterol fumarate), and even more preferably albuterol (albuterol sulfate).

[0015] Albuterol (also known as salbutamol) is typically administered as a sulfate and is a preferred rescue medication in the present invention.

[0016] Anticholinergic agents (antimuscarinic agents) block the neurotransmitter acetylcholine by selectively blocking its receptors within nerve cells. Upon topical application, anticholinergic agents act mainly on M 3 muscarinic receptors located within the airways, causing smooth muscle relaxation and bronchodilation. Long-acting muscarinic antagonists (LAMA) include, for example, tiotropium (tiotropium bromide), oxitropium (oxitropium bromide), acridinium (acridinium bromide), ipratropium (ipratropium bromide), glycopyrronium (glycopyrronium bromide), oxybutynin (oxybutynin hydrochloride or oxybutynin hydrobromide), tolterodine (tolterodine tartrate), trospium (trospium chloride), solifenacin (solifenacin succinate), fesoterodine (fesoterodine fumarate), and darifenacin (darifenacin hydrobromide).

[0017] To date, numerous approaches have been made in preparing and prescribing these drugs for administration by inhalation using dry powder inhalers (DPIs), pressurized metered-dose inhalers (pMDIs), nebulizers, and the like.

[0018] According to the guidelines of GINA (Global Initiative for Asthma: International Guidelines for Asthma Management), a stepwise approach is taken for the treatment of asthma. In the first step, representing mild asthma, patients are given a SABA as needed, such as albuterol sulfate. Patients may also be given a low-dose ICS-formoterol or a low-dose ICS as needed whenever a SABA is prescribed. In the second step, a regular low-dose ICS is given together with a SABA or a low-dose ICS-formoterol as needed. In the third step, a LABA is added. In the fourth step, the dose of each drug is increased, and in the fifth step, additional treatments, such as an anticholinergic drug or a low-dose oral corticosteroid, are included. That is, each step can be regarded as a treatment plan, and the treatment is configured according to the degree of acute severity of the respiratory disease, respectively.

[0019] COPD is a major cause of death worldwide. COPD is a heterogeneous long-term disease including chronic bronchitis, emphysema, and further airway narrowing. The lesions in patients with COPD are mainly limited to the airways, lung parenchyma, and pulmonary vasculature. These lesions reduce the healthy ability to inhale and exhale gases in the lungs.

[0020] Bronchitis is characterized by inflammation of the trachea over a long period. Common symptoms include wheezing, dyspnea, cough, and expectoration of sputum, all of which are troublesome and harmful to the patient's quality of life. Emphysema is also associated with long-term bronchial inflammation, in which case, due to the body's reaction to the inflammation, destruction of lung tissue and progressive airway narrowing occur. Eventually, the lung tissue loses its natural elasticity and expands. As a result, the gas exchange function deteriorates, and the inhaled gas is often trapped in the lungs. This causes localized hypoxia and reduces the amount of oxygen supplied to the patient's bloodstream per inhalation. And the patient experiences dyspnea.

[0021] Patients living with COPD experience their various symptoms, some of which may be daily. The severity of the disease is determined by various factors, but most commonly, it correlates with the progression of the disease. Their symptoms represent stable COPD regardless of the severity of the disease, and this condition is maintained and controlled through the administration of various medications. The treatment methods can be changed, but often, inhaled bronchodilators, anticholinergic drugs, long-acting and short-acting β 2 agonists, and inhaled corticosteroids are included. The medications are often administered as monotherapy or combination therapy.

[0022] Patients are classified according to the severity of their COPD using the categories defined in the GOLD (Global Initiative for Chronic Obstructive Lung Disease, Inc.: Global Initiative for Chronic Obstructive Lung Disease) guidelines. The categories consist of A to D, and the recommended first-choice treatment options vary for each category. Patient group A is prescribed a short-acting muscarinic antagonist (SAMA) or a short-acting β 2 agonist (SABA) as needed. Patient group B is prescribed a long-acting muscarinic antagonist (LAMA) or a long-acting β 2A stimulant drug (LABA) is prescribed. Patient group C is prescribed an inhaled corticosteroid (ICS) + LABA, or a LAMA. Patient group D is prescribed an ICS + LABA and / or a LAMA.

[0023] Patients suffering from respiratory diseases such as asthma or COPD suffer from periodic exacerbations beyond the range of daily basic condition changes. An exacerbation is an acute deterioration of the condition that requires additional treatment, that is, treatment beyond the scope of maintenance treatment.

[0024] For asthma, additional treatment for moderate exacerbations consists of repeated administration of SABA and oral corticosteroids and / or controlled oxygen inhalation (the latter requires hospitalization). For severe exacerbations, an anticholinergic agent (typically ipratropium bromide), nebulized SABA, or magnesium sulfate is added.

[0025] For COPD, additional treatment for moderate exacerbations consists of repeated administration of SABA, oral corticosteroids, and / or antibiotics. For severe exacerbations, controlled oxygen inhalation and / or respiratory support (both of which require hospitalization) are added.

[0026] In this specification, the term "exacerbation" includes both moderate exacerbations and severe exacerbations.

[0027] The present invention relates to a treatment approach that can intervene early in the treatment of patients by predicting exacerbations of respiratory diseases, thereby improving the prognosis of the patients.

[0028] According to the present invention, a system for calculating the probability of exacerbation of COPD in a subject is provided. The system includes a first inhaler for administering a rescue medication to the subject. The rescue medication is suitable for treating the worsening of respiratory symptoms, for example, by rapidly dilating the bronchi and bronchioles upon inhalation of the medication. The first inhaler has a usage detection system configured to measure rescue inhalations performed by the subject using the first inhaler. The system optionally has a second inhaler for administering a maintenance medication to the subject during regular inhalations. The sensor system is configured to measure parameters related to the airflow during rescue inhalations and / or during regular inhalations when the second inhaler is used in the system.

[0029] The rescue medication is as described above and typically consists of a SABA or a fast-acting LABA such as formoterol (formoterol fumarate). The rescue medication also has, for example, the form of a combination product such as ICS-formoterol (formoterol fumarate), typically budesonide-formoterol (budesonide-formoterol fumarate). Such an approach is called "MART" (maintenance and rescue therapy). However, the presence of the rescue medication, as represented herein, means that it is the first inhaler. This is because the presence of the rescue medication is decisive in the weighted model. Thus, it covers both the rescue medication and the combination of the rescue medication and the maintenance medication. In contrast, the second inhaler, when used, is only used in the maintenance aspect of the treatment and not for rescue purposes. The important difference is that the first inhaler is used as needed, while the second inhaler is used regularly, at a given time.

[0030] The system further comprises a processor configured to measure the number of rescue inhalations within a first period and receive measurement values of parameters during at least several rescue and / or regular inhalations. The processor calculates the probability of COPD exacerbation based on the measurement values of the parameters during rescue inhalations using a weighted model. The model is weighted such that the measurement values of the parameters are more significant than the number of rescue inhalations in the calculation of the probability. According to the present invention, there is further provided a method for calculating the probability of COPD exacerbation in a subject. The method uses a weighted model. Any of the preferred embodiments described with respect to the system of the present invention may be applied to this method, and vice versa.

[0031] Attempts have been made to evaluate the risk of exacerbation of acute respiratory diseases such as asthma or COPD by monitoring factors and environmental factors related to various subjects. Challenges have been made regarding which factors should be taken into account and which factors should be ignored. By ignoring factors that have little or negligible impact on the risk assessment, it is possible to calculate the risk more efficiently using resources for calculations such as fewer processing resources, battery power, and memory capacity. More importantly, it is to improve the calculation accuracy of the probability of exacerbation of acute respiratory diseases. A more accurate risk assessment can realize a more effective alarm system, and as a result, medical intervention can be more appropriately performed on the subject. That is, a more accurate assessment of the risk of exacerbation has the potential to guide intervention on the subject in case of an emergency risk.

[0032] When the probability of exacerbation is high, a step - by - step change in the treatment plan can be justified, for example, for the subject, with respect to the treatment that was configured for the subject in case of an emergency risk. Or, when the probability of exacerbation is low over a long period, a more accurate probability calculation can be used as a guideline to justify downgrading or withdrawing the current treatment method. This means, for example, that the subject is not required to take a high - dose drug that is no longer in balance with their respiratory disease condition.

[0033] The present inventor has conducted extensive clinical research and, as will be described in detail below, by using a weighted model that bases the calculation of the probability of exacerbation of COPD on both the number of emergency inhalations of rescue medications performed by the subject within a certain period (the first period) and the parameter (measurement value) related to the airflow between inhalations of rescue and / or maintenance medications, it has been found that the accuracy of the calculation of the probability of exacerbation of COPD can be improved.

[0034] The first period corresponds to the sampling period during which the number of emergency inhalations is measured. The first period is, for example, from 1 to 30 days. This sampling period is selected so that the period is of an appropriate length for data collection of the number of emergency inhalations. If the sampling period is too short, an insufficient amount of data necessary for a reliable prediction of exacerbation cannot be collected, and if the sampling period is too long, an averaging effect occurs that makes it difficult to distinguish a short - term trend that is meaningful for diagnosis or prediction.

[0035] By using both the number of emergency inhalations and the parameter (value), for example, a model can be formed that enables a prediction with higher accuracy than a model that ignores either of these factors. Furthermore, from clinical research, it has been found that the parameter related to the airflow during inhalation and the trend regarding the parameter are more significant in the probability calculation than the number of emergency inhalations. The number of emergency inhalations is still a significant factor in the calculation of the probability of exacerbation, but compared to the parameter, the influence on the probability is small. Therefore, the accuracy of the probability calculation is further improved by weighting the model such that the parameter is more significant than the number of rescue inhalations in the probability calculation.

[0036] The model has, for example, a first weighting coefficient associated with the parameter and a second weighting coefficient associated with the number of inhalations. When standardized to evaluate the different units used to quantify the number of rescue inhalations (or the tendency related to the use of rescue medications) and the parameter, the first weighting coefficient becomes larger than the second weighting coefficient, thereby ensuring that the parameter is more significant than the number of rescue inhalations in the probability calculation.

[0037] The probability calculation is based on parameters (values) related to the airflow during rescue inhalations and / or during regular inhalations using a second inhaler, if a second inhaler is used. The parameter corresponds to a single factor related to the airflow during inhalation or may include a plurality of such factors. For example, the parameter consists of at least one of the maximum inhalation flow rate, inhalation volume, inhalation duration, and inhalation speed. The time versus the maximum inhalation flow rate gives, for example, a measured value of the inhalation speed.

[0038] Placing the basis of the probability calculation on the parameter means that the model uses one or more factors related to the airflow during inhalation and / or one or more tendencies related to each of the factors. Such tendencies correspond to variations in each factor.

[0039] The first weighting coefficient weights one or more factors related to the airflow during inhalation and / or one or more tendencies related to each of the factors.

[0040] More generally, parameters related to the airflow during rescue inhalations and / or during regular inhalations (including any related tendencies, for example) have a significance / importance in the model of 55% - 95%, preferably 65% - 90%, most preferably 75% - 95%, for example 80% (relative to other factors).

[0041] The probability calculation is also based in part on the number of rescue inhalations. Performing the probability calculation based on the number of rescue inhalations means that the model uses the absolute number of rescue inhalations and / or a trend based on the number of rescue inhalations within the first period. Such a trend consists of the variation in the number of rescue inhalations rather than the number of rescue inhalations as needed.

[0042] The second weighting factor weights one or more trends based on the absolute number of rescue inhalations and / or the number of rescue inhalations.

[0043] The trend based on the number of rescue inhalations includes, for example, the number of inhalations performed within a specific time of day. Thus, the number of night-time inhalations is included as one factor, for example, in the number of inhalations.

[0044] More generally, the number of rescue inhalations (including any relevant trends, for example) has a significance / importance in the model of 2% to 30%, preferably 5% to 25%, most preferably 10% to 20%, for example 15%.

[0045] The probability of a COPD exacerbation is the probability of an impending COPD exacerbation occurring within an exacerbation period following the first period. Thus, the model enables the calculation of the probability of a COPD exacerbation occurring within a pre-determined period called the "exacerbation period" following the first period during which inhalation data, i.e., the number of rescue inhalations, and parameter data are collected. The exacerbation period is, for example, 1 to 10 days, for example 5 days. The exacerbation period is selected based on the model's ability to predict an exacerbation within that period, ensuring that the pre-determined period is long enough to take appropriate treatment steps if necessary.

[0046] In some embodiments, parameters related to a living body are included in the model to improve the accuracy of the weighted model. In such embodiments, the processor is configured to receive, for example, parameters (parameter values) related to the living body. A data input unit is included in the system, for example, so that the subject and / or medical personnel can input parameters related to the living body.

[0047] The model is weighted such that, for example, parameters related to the living body have a lower significance in probability calculations than parameters related to the airflow during inhalation. In other words, a third weighting coefficient is associated with the parameters related to the living body, and the third weighting coefficient is set to be smaller than the first weighting coefficient associated with the parameters related to the airflow. The third weighting coefficient is set to be greater than or smaller than the second weighting coefficient associated with the number of rescue inhalations.

[0048] Preferably, the third weighting coefficient is smaller than the second weighting coefficient. In order of predictive power, the parameters related to the airflow during inhalation have the greatest influence, followed by the number of rescue inhalations, and then the parameters related to the living body.

[0049] Parameters related to the living body include, for example, one or more parameters selected from weight, height, body mass index, blood pressure including systolic and / or diastolic blood pressure, gender, race, age, smoking history, sleep / activity pattern, history of exacerbation, other treatments administered to the subject or other drugs administered, and so on. In a preferred embodiment, the parameters related to the living body include age, body mass index, and history of exacerbation.

[0050] More generally, the parameters related to the living body have a significance / importance (e.g., weight) of 1% - 12% in the model, preferably 3% - 10%, most preferably 4% - 6%, for example 5%.

[0051] Additionally, additional data sources such as environmental data regarding weather or pollution levels can be added to the model. Such additional data is weighted so as to have a lower significance with respect to probability calculations than the data of the inhalation parameters and, optionally, also lower significance than the data of the number of rescue inhalations.

[0052] The number of maintenance / regular inhalations provides useful information for predicting exacerbation, either selectively or additionally. This is because a lower number of maintenance / regular inhalations (indicating insufficient compliance with the dosing of maintenance medications) increases the risk of exacerbation.

[0053] In a relatively simple example, an "increase in the number of rescue inhalations using a first inhaler" and / or a "decrease in the number of regular inhalations using a second inhaler" (indicating low compliance with the treatment plan) for a reference period regarding the subject in question, together with the inhalation parameters, represent a deterioration of lung function that increases the probability of exacerbation of the lung disease.

[0054] In a specific example, when the compliance with the dosing of the maintenance medication decreases from 80% to 55%, the use of the rescue inhaler increases by 67.5%, the maximum inspiratory flow rate decreases by 34%, and the inhalation volume decreases by 23% (all of which are the rates of change from the patient's reference values), and there were 2 exacerbations in the previous year and the BMI exceeds 28, then using the ROC-AUC (see the explanations regarding Figures 12 and 17 below), the probability of exacerbation in the next 5 days is 0.87.

[0055] The model is a linear or non-linear model. The model consists of, for example, a machine learning model. For example, a supervised model such as a supervised machine learning model is used. Regardless of the specific form of the model used, as already explained, the model is more sensitive to the inhalation parameters than the number of rescue inhalations, that is, it is configured to respond. This sensitivity corresponds to the "weighting" of the weighted model.

[0056] In a non-limiting example, the model is configured to use a decision tree. Other suitable techniques such as neural networks or deep learning models may also be considered by those skilled in the art.

[0057] Irrespective of predicting the exacerbation of a respiratory disease, the system's processor calculates the probability of exacerbation based on the number of inhalations, inhalation parameters, and an indicator related to the condition of the respiratory disease from which the subject is suffering. By including the indicator in the prediction, the prediction accuracy is improved. The reason is that the user-entered indicator helps to validate and make more certain the predicted value of the probability assessment derived by considering the number of inhalations and inhalation parameters without using, for example, the user-entered indicator.

[0058] In one embodiment, the processor calculates an initial probability of exacerbation of a respiratory disease based on the recorded number of inhalations and the received inhalation parameters (values), but not based on the indicator. The initial probability is calculated, for example, using a weighted model as described above. And the probability, i.e., the overall probability, is calculated based on the number of inhalations, inhalation parameters, and the received indicator related to the symptoms of the respiratory disease from which the subject is suffering. For example, the overall probability is calculated based on the initial probability and the received indicator.

[0059] The initial probability determines, for example, the risk of exacerbation in the next 10 days. The overall probability taking into account the indicator of the condition of the respiratory disease from which the subject is suffering determines, for example, the risk of exacerbation in the next 5 days. That is, by including the indicator in the probability calculation, a more accurate prediction in a shorter period is possible.

[0060] By including the user-entered indicator in the probability calculation, one or more of the positive and negative predictive values, the sensitivity of the prediction, i.e., the ability of the system / method to correctly determine that the subject is at risk (true positive rate), and the specificity of the prediction, i.e., the ability of the system / method to correctly determine that the subject is not at risk (true negative rate) are enhanced.

[0061] Data on the number of inhalations and inhalation parameters represent, for example, the deviation from the reference value of the subject over a 10-day period prior to exacerbation. By including the user-entered indicators in subsequent predictions, the positive and negative predictive values, and the sensitivity and specificity of the prediction system / method are improved.

[0062] The processor is configured to control, for example, a user interface that outputs a prompt to the user and receives input of an indicator from the user. The prompt is output based on an initial probability calculated based on the number of inhalations and inhalation parameters, but not based on the indicator. The prompt is output, for example, based on an initial probability that has reached or exceeded a predetermined threshold. In this way, the user is prompted by the system to enter an indicator based on the initial probability indicating the likelihood of an impending exacerbation. When the user enters an indicator, the (overall) probability considering the indicator helps to support and justify the initial probability.

[0063] This is regarded as, for example, "analysis-driven use" of the indicator. User input is required when data on the number of inhalations and inhalation parameters indicate a possible exacerbation of the subject's respiratory disease.

[0064] The user interface prompts the user or subject to enter an indicator, for example, by completing a short questionnaire through a pop-up notification. The logic for determining when to issue the pop-up notification is triggered by deviations in key variables such as the number and / or duration of rescue and / or maintenance inhalations, and changes in inhalation parameters.

[0065] Optionally or additionally, the system is configured to receive an indicator when the user selects to input the indicator through the user interface. For example, when a healthcare provider determines that the indicator is useful for making the calculation of the initial probability certain. This is regarded as, for example, a "demand-driven" use of the indicator. This demand is made by the patient or the physician in charge of the patient, for example, prior to or during the evaluation by a medical professional.

[0066] Thus, the user is only prompted to input the indicator when this is considered necessary by the system and / or the healthcare provider. This advantageously reduces the burden on the subject and enables the subject to input the indicator when requested or prompted to do so, i.e., when such input is desirable for monitoring the subject's respiratory disease. Inputting the indicator in these examples is more feasible than when the subject is routinely prompted to input the indicator.

[0067] In one embodiment, the user interface is configured to provide a plurality of selectable respiratory disease conditions for the user to choose from. In this case, the indicator is determined by the user's selection of at least one of the disease condition options.

[0068] For example, the user interface displays a questionnaire consisting of a plurality of questions, each answer corresponding to an indicator. The user, such as the subject or the healthcare provider in charge of the subject, uses the user interface to input answers to the questions.

[0069] The questionnaire is relatively short, i.e., consists of a relatively small number of questions, in order to minimize the burden on the subject. Nevertheless, the number and content of the questions are set such that the calculation of the probability of exacerbation by the indicator is more certain than when no indicator is input at all.

[0070] More generally, the purpose of the questionnaire is to check current or relatively recent (e.g., within the last 24 hours) indicators in order to immediately understand the subject's living conditions (related to respiratory diseases) based on a small number of questions that can be answered relatively quickly. The questionnaire is translated into the local language of the subject.

[0071] General control questionnaires, especially the ACQ / T (Asthma Control Questionnaire / Test) for asthma or the CAT (COPD Assessment Test) for COPD, tend to focus on recalling the patient's past medical conditions. Focusing on the past rather than the present may negatively affect the value for the purpose of predictive analysis.

[0072] Next, a non-limiting example of such a questionnaire will be described. For each question, the subject selects one option from the options indicating the situation: always (5), usually (4), sometimes (3), rarely (2), and never (1). 1. How frequently do you experience shortness of breath, or what is the rate at which your shortness of breath occurs? 2. How frequently do you experience coughing, or what is the rate at which your coughing occurs? 3. How frequently do you experience wheezing, or what is the rate at which your wheezing occurs? 4. How frequently do you experience chest tightness, or what is the rate at which your chest tightness occurs? 5. How frequently do you experience nocturnal attacks / effects on sleep, or what is the rate of your nocturnal attacks / effects on sleep? 6. How frequently do you experience restrictions at work, school, or home, or what is the rate of your restrictions at work, school, or home? Another example of the questionnaire is as follows. 1. Are you experiencing more respiratory symptoms than usual (yes / no)? If the answer is "yes": 2. Have you experienced more chest pressure or shortness of breath (Yes / No)? 3. Have you experienced more coughing (Yes / No)? 4. Have you experienced more wheezing (Yes / No)? 5. Is it affecting your sleep (Yes / No)? 6. Is it restricting your activities at home / work / school (Yes / No)?

[0073] The answers to these questions are used, for example, to calculate a score, which is included in or corresponds to an indicator of the condition of the respiratory disease suffered by the subject.

[0074] In one embodiment, the user interface is configured to provide selectable icons, such as emoji - type icons and / or check - boxes and / or sliders and / or dials, as situation - indicating options. Thus, the user interface provides a direct and intuitive way for the patient to input an indicator of the condition of the respiratory disease suffered. Such an intuitive input method is particularly advantageous when the subject himself / herself inputs indicators regarding himself / herself. Because user input that can be performed relatively easily is hardly hindered even by the deterioration of the subject's respiratory disease.

[0075] A suitable user interface is used for the purpose of enabling user input of an indicator of the condition of the respiratory disease suffered by the subject. For example, the user interface comprises or consists of the (first) user interface of a user - owned device. The user - owned device is, for example, a personal computer, a tablet computer and / or a smartphone. When the user - owned device is a smartphone, the (first) user interface corresponds to, for example, the touch screen of the smartphone.

[0076] 1. In an embodiment, the system's processor includes at least in part a (first) processor within a user-owned device. Optionally or additionally, the first inhaler and / or the second inhaler has, for example, a (second) processor, and the system's processor includes at least in part the (second processor) included in the inhaler.

[0077] The method of the present invention is provided for calculating the probability of exacerbation of COPD in a subject. The method measures the number of rescue inhalations of a rescue medicine suitable for the treatment of the subject's impending respiratory disease performed by the subject within a first period, measures parameters related to the airflow during at least several rescue inhalations and / or during the regular inhalation of a maintenance medicine performed by the subject, and uses a weighted model to calculate the probability of exacerbation of COPD based on the number of rescue inhalations and the measured values of the parameters, and the model is weighted such that the measured values of the parameters are more significant than the number of rescue inhalations in the calculation of the probability.

[0078] Furthermore, according to the present invention, there is provided a method for treating a subject's COPD, which comprises implementing the method for calculating the probability of exacerbation of COPD described above, determining whether the probability reaches or exceeds a predetermined upper limit, or alternatively, determining whether the probability reaches or is lower than a predetermined lower limit, and treating COPD based on the probability that has reached or exceeded the upper limit, or based on the probability that has reached or is lower than the lower limit.

[0079] This treatment method changes the existing treatment method. The existing treatment method has a first treatment plan, and the change of the existing treatment method consists of changing from the first treatment plan to a second treatment plan based on the probability that has reached or exceeded a predetermined threshold, and the second treatment plan is configured for exacerbation of COPD having a higher risk than the first treatment plan.

[0080] More accurate risk calculations using the weighted model facilitate the realization of a more effective warning system, which in turn leads to appropriate medical intervention for the subject. That is, a more accurate assessment of the risk of exacerbation makes it possible to guide intervention for the subject in the event of a significant risk. In particular, the intervention includes implementing a second treatment plan. This includes, for example, advancing the subject to a higher stage as defined by the GINA or GOLD guidelines. Such proactive intervention means that the subject does not have to continue to suffer from exacerbation or be exposed to the associated risks in order to proceed to a second treatment plan that should be justified.

[0081] In one embodiment, the second treatment plan consists of administering a biopharmaceutical to the subject. Relatively expensive biopharmaceuticals mean that stepping up the subject's treatment to include the administration of biopharmaceuticals often requires careful consideration and justification. The system and method according to the present invention can justify the administration of a biopharmaceutical by providing a reliable prediction regarding the risk that the subject will experience an exacerbation. For example, when the calculated probability reaches or exceeds a pre-determined minimum number and an upper limit representing a high-risk exacerbation, the administration of the biopharmaceutical is quantitatively justified, and the biopharmaceutical is administered accordingly.

[0082] More generally, the biopharmaceutical consists of any one or a combination of two or more of omalizumab, mepolizumab, reslizumab, benralizumab, and dupilumab.

[0083] Changing the existing COPD treatment method consists of changing from the first treatment plan to a third treatment plan created for exacerbation of a respiratory disease with a lower risk than the first treatment plan based on a probability that reaches or is lower than a pre-determined lower limit.

[0084] For example, when the probability of exacerbation is low over a relatively long period, the probabilistic calculation of certain accuracy is used as advice to justify a downgrading or even exclusion of an existing treatment plan. In particular, the subject is transferred from the first treatment plan to a third treatment plan created for exacerbation of respiratory relaxation with a lower risk than the first treatment plan. This includes, for example, advancing the subject to a lower stage as defined in the GINA or GOLD guidelines.

[0085] Also, according to the present invention, there is provided a method for diagnosing exacerbation of COPD, which comprises implementing a method for calculating the probability of exacerbation of COPD in the subject described above, determining whether the probability reaches or exceeds a predetermined upper limit representing exacerbation of COPD, and diagnosing exacerbation of COPD based on the probability that has reached or exceeded the upper limit.

[0086] Furthermore, according to the present invention, there is provided a method for diagnosing the severity of COPD in a subject, which comprises implementing a method for calculating the probability of exacerbation of COPD in the subject described above, determining whether the probability has reached or exceeded a predetermined upper limit indicating that COPD is more severe, or determining whether the probability has reached or fallen below a lower limit indicating that COPD is less severe, diagnosing that COPD has become more severe based on the probability that has reached or exceeded the upper limit, or diagnosing that COPD has become less severe based on the probability that has reached or fallen below the lower limit.

[0087] Furthermore, a method for delineating a subpopulation of subjects is provided, the method comprising implementing the method defined above for each subject in a population of subjects, thereby calculating the probability of exacerbation of COPD for each subject in the population, providing a threshold probability or range of thresholds that distinguish the probabilities calculated for the subpopulation from the probabilities calculated for the remainder of the population, and using the threshold probability or range of thresholds to delineate the subpopulation from the remainder of the population.

[0088] Figure 1 is a block diagram of a system 10 according to an embodiment of the present invention. The system 10 includes a first inhaler 100 and a processor 14. The first inhaler 100 is used to administer an emergency drug such as a SABA to a subject. The SABA includes, for example, albuterol. The first inhaler 100 has a sensor system 12A and / or a usage detection system 12B.

[0089] The system 10 is optionally referred to as an "inhaler assembly", for example.

[0090] The first inhaler is optionally referred to as an "emergency inhaler", for example.

[0091] The second inhaler is optionally referred to as a "maintenance inhaler" or a "management inhaler", for example.

[0092] The number of emergency inhalations is measured by the usage detection system 12B included in the first inhaler 100.

[0093] The sensor system 12A is configured to measure parameters. The sensor system 12A has one or more sensors such as, for example, one or more pressure sensors and / or temperature sensors and / or humidity sensors and / or direction sensors and / or acoustic sensors and / or optical sensors. The pressure sensors include barometric pressure sensors (e.g., atmospheric pressure sensors), differential pressure sensors, absolute pressure sensors, and the like. The sensors use microelectromechanical system (MEMS) and / or nanoelectromechanical system (NEMS) technologies.

[0094] The pressure sensor is particularly suitable for measuring parameters, because it is monitored by measuring the pressure changes related to the airflow during inhalation by the subject. As will be described in more detail below with reference to FIGS. 18 to 22, the pressure sensor is arranged, for example, within or in communication with the flow path through which air and medicament are inhaled by the subject during inhalation. Another method of measuring parameters, such as using a flow sensor, will also be apparent to those skilled in the art.

[0095] Optionally or additionally, the sensor system 12A comprises a differential pressure sensor. The differential pressure sensor consists of, for example, a dual-port type sensor for measuring the pressure difference across a part of the air flow path through which the subject inhales. Alternatively, a single-port gauge type sensor is used. The latter measures the pressure difference in the air flow path during inhalation and when there is no air flow. The difference in the measured values corresponds to the pressure drop related to inhalation.

[0096] Although not shown in FIG. 1, the system 10 further comprises a second inhaler for administering maintenance medication to the subject during regular inhalation. The second inhaler has a sensor system 12A and / or a usage detection system 12B that are different from those of the sensor system 12A and / or the usage detection system 12B of the first inhaler 100. The sensor system 12A of the second inhaler is configured to measure parameters during regular inhalation. For example, the sensor system 12A has another pressure sensor, such as another microelectromechanical system pressure sensor or another nanoelectromechanical system pressure sensor, for measuring parameters during the inhalation of maintenance medication.

[0097] Thus, the inhalation of one or both of the emergency medication and the maintenance medication is used to collect information regarding the lung function and / or the health of the lungs of the subject. When the first and second inhalers are used, the prediction accuracy of impending exacerbation is improved by the additional inhalation data provided by monitoring both regular inhalation and emergency inhalation.

[0098] Regular inhalation and emergency inhalation each reduce the pressure in the air flow path compared to when inhalation is not taking place. The point at which the pressure is at its lowest corresponds to the maximum inhalation flow rate. Sensor system 12A detects this point during inhalation. The maximum inhalation flow rate varies from inhalation to inhalation and depends on the subject's medical condition. A lower maximum inhalation flow rate is recorded, for example, when the subject is approaching a deterioration. As used herein, the term "minimum maximum inhalation flow rate" means the minimum of the maximum inhalation flow rates recorded for inhalations performed using the first and / or second inhaler within a defined (second) period.

[0099] The pressure change associated with each inhalation is used selectively or additionally to measure the inhalation volume. This is achieved, for example, by measuring the flow rate over the course of an inhalation using the pressure change measured by sensor system 12A during the inhalation and then deriving the total inhalation volume. A lower inhalation volume is recorded, for example, when the subject is approaching a deterioration. This is because the subject's ability to inhale decreases. As used herein, the term "minimum inhalation volume" means the minimum inhalation volume recorded for inhalations performed using the first and / or second inhaler within a defined (third) period.

[0100] The pressure change associated with each inhalation is used selectively or additionally to measure the inhalation duration. For example, the time from the first pressure drop measured by pressure sensor 12A that coincides with the start of inhalation until the pressure returns to the pressure when no inhalation is taking place is recorded. A shorter inhalation duration is recorded, for example, when the subject is approaching a deterioration. This is because the subject's ability to inhale for a long time decreases. As used herein, the term "minimum inhalation duration" means the shortest inhalation duration recorded for inhalations performed using the first and / or second inhaler within a defined (fourth) period.

[0101] In one embodiment, the parameter includes, instead of or in addition to, for example, the maximum inhalation flow rate and / or the inhalation volume and / or the inhalation duration, the time versus the maximum inhalation flow rate. This time versus the maximum inhalation flow rate is recorded, for example, from the first pressure drop measured by the sensor system 12A that coincides with the start of inhalation until the pressure reaches the minimum value corresponding to the maximum flow rate. Subjects at greater risk of exacerbation achieve the maximum inhalation flow rate over a longer period of time.

[0102] In a non-limiting example, the first and / or second inhaler is configured such that, for normal inhalation, each drug is administered in about 0.5 seconds following the start of inhalation. The inhalation of a subject who reaches the maximum inhalation flow rate after the elapse of the 0.5 seconds, such as after about 1.5 seconds, represents a partially impending exacerbation.

[0103] The usage detection system 12B is configured to record an inhalation by a subject (for example, each emergency inhalation by the subject when the inhaler is an emergency inhaler, or each maintenance inhalation by the subject when the inhaler is a maintenance inhaler). In a non-limiting example, the first inhaler 100 includes a drug tank (not shown in FIG. 1) and a metering assembly (not shown in FIG. 1) configured to meter a single dose of the drug from the drug tank. The usage detection system 12B is configured to record the metering of a single dose of the drug by the metering assembly, and each metering represents an emergency inhalation performed by the subject using the first inhaler 100. Thus, the inhaler 100 is configured to monitor the number of emergency inhalations of the drug. This is because a single dose of the drug must be metered by the metering assembly before being inhaled by the subject. One non-limiting example of the configuration for metering will be described in more detail with reference to FIGS. 18-22.

[0104] Optionally or additionally, the usage detection system 12B registers each inhalation in various ways and / or based on additional or alternative feedback that would be apparent to those skilled in the art. For example, the usage detection system 12B is configured to register an inhalation by a subject when feedback from the sensor system 12A indicates that an inhalation by the user is occurring (e.g., when a pressure measurement or flow rate exceeds a predetermined threshold associated with a successful inhalation). Further, in some examples, the usage detection system 12B is configured to register an inhalation when an inhaler switch or user input to an external device (e.g., a touch screen of a smartphone) is manually performed by the subject before, during, or after an inhalation.

[0105] A sensor (e.g., a pressure sensor) is included in the usage detection system 12B, for example, to register each inhalation. In such an example, the usage detection system 12B and the sensor system 12A use respective sensors (e.g., pressure sensors) or a common sensor (e.g., a common pressure sensor) configured to perform both the function of detecting usage and the function of detecting inhalation parameters.

[0106] When a sensor is included in the usage detection system 12B, the sensor is used, for example, to confirm that a single dose of a medicament metered by the metering assembly has been inhaled by the user or to evaluate the extent of the inhalation, as described in more detail with reference to FIGS. 18-22.

[0107] In one embodiment, the sensor system 12A and / or the usage detection system 12B includes an acoustic sensor. The acoustic sensor in this embodiment is configured to detect noise generated when a subject performs an inhalation using each inhaler. The acoustic sensor consists of, for example, a microphone.

[0108] In a non-limiting example, each inhaler has a capsule arranged to rotate when the subject inhales using the device. The rotation of the capsule generates noise that can be detected by an acoustic sensor. That is, the rotation of the capsule provides appropriately interpretable noise, for example, to lead to data on use and / or inhalation parameters such as rattling sounds.

[0109] An algorithm is used, for example, to interpret acoustic data to measure data on use (when the acoustic sensor is included in the use detection system 12B) and / or to measure parameters regarding the airflow during inhalation (when the acoustic sensor is included in the sensor system 12A).

[0110] For example, an algorithm such as that described in Colthorpe et al., "Adding Electronics to the Breezhler; Satisfying the Needs of Patients", Respiratory Drug Delivery, 2018, pages 71 - 79 is used. Once the generated sound is detected, the algorithm processes the raw acoustic data and generates data on use and / or inhalation parameters.

[0111] The processor 14 included in the system 10 measures the number of rescue and / or inhalation events during a first period and receives the parameters measured for each of the rescue and / or regular inhalations. In FIG. 1, as indicated by the arrows between the sensor system 12A and the processor 14, and between the use detection system 12B and the processor 14, the processor receives data on inhalation and parameters from the use detection system 12B and the sensor system 12A respectively. The processor 14 is further configured to calculate the probability of exacerbation of a respiratory disease based on the number of rescue inhalations and the parameters, using a weighted model, as will be described in more detail with reference to FIGS. 3 - 17.

[0112] In a non-limiting example, the processor 14 is arranged independently from each of the first and / or second inhalers. In this case, the processor 14 receives data on the number of rescue inhalations and parameters transmitted from the sensor systems 12A and the usage detection system 12B of the first and / or second inhalers. By processing the data in an external processing unit such as a processing unit of an external device, the battery life of the inhaler is advantageously maintained.

[0113] In another non-limiting example, the processor 14 is an essential part of the first and / or second inhaler and is housed, for example, within the main housing (not shown in FIG. 1) or the upper cap of the first and / or second inhaler. In such an example, there is no need to rely on a connection to an external device. This is because the calculation of the probability of exacerbation of the respiratory disease is performed only inside the first and / or second inhaler. The first and / or second inhaler includes a user interface, such as a light source, a screen, a loudspeaker, etc., for informing the subject of the result of the probability calculation. In some examples, rather than informing the probability numerically, more intuitive means of informing the subject of the risk are used, such as using light sources of different colors according to the calculated value of the probability. That is, the first and / or second inhaler, for example, takes proactive steps such as the subject inhaling the rescue medicine one or more times, and prompts the subject to reduce or eliminate the risk of exacerbation.

[0114] It is also possible for some of the functions of the processor 14 to be executed by an internal processing unit included in the first and / or second inhaler, and for other functions of the processor, such as the probability calculation itself, to be executed by an external processing unit.

[0115] More generally, system 10 has a communication module (not shown in FIG. 1) configured to notify, for example, a calculated probability to a subject and / or a healthcare worker such as a physician. Then, the subject and / or the physician take appropriate steps based on the calculated probability of exacerbation of the respiratory disease. For example, when a smartphone processing unit is included in the processor, communication functions of the smartphone such as SMS, e-mail, Bluetooth (registered trademark), etc. are used to notify the calculated probability to the healthcare worker.

[0116] FIG. 2 shows a non-limiting example of system 10 that calculates the probability of exacerbation of a respiratory disease in a subject. An optionally weighted model, optionally called a respiratory disease exacerbation risk prediction model, is used to calculate the probability, and the calculation result is provided to the subject and / or caregiver and / or healthcare worker.

[0117] System 10 has a first inhaler 100, an external device 15 (e.g., a mobile device), a public and / or private network 16 (e.g., the Internet, a cloud network, etc.) and a personal data storage device 17. The external device 15 includes, for example, a smartphone, a personal computer, a laptop computer, a wireless media device, a media streaming device, a tablet device, a wearable device, a Wi-Fi or wireless communication-capable television, or other suitable device connectable via an Internet protocol. For example, the external device 15 is configured to transmit and / or receive RF signals through a Wi-Fi communication link, a Wi-MAX communication link, a Bluetooth or Bluetooth Smart communication link, a Near Field Communication (NFC) link, a cellular phone communication link, or a Television White Space (TVWS) communication link or a combination of two or more thereof. The external device 15 transmits data to the personal data storage device 17 through the public and / or private network 16.

[0118] The first inhaler 100 has a communication circuit such as a Bluetooth radio for transmitting data to an external device 15. The data includes the data on the number of inhalations and parameters described above.

[0119] The first inhaler 100 also receives data such as program instructions, changes to the operating system, information regarding a single dose, warnings or notifications, approvals, etc. from the external device 15, for example.

[0120] The external device 15 includes at least a part of the processor 14, thereby processing and analyzing the data on the number of inhalations and parameters. For example, as represented by block 18A, the external device 15 processes data such as data for calculating the probability of exacerbation of a respiratory disease and provides such information to the personal data storage device 17 for automatic storage.

[0121] In some non-limiting examples, the external device 15 also identifies no-inhalation events and / or low-inhalation events and / or good-inhalation events and / or over-inhalation events and / or exhalation events by processing the data, as represented by block 18B. The external device 15 also identifies inadequate use, excessive use, and optimal use by processing the data, as represented by block 18C. The external device 15 processes the data to estimate, for example, the number of doses administered and / or remaining, and identifies an error state such as an error state related to a timestamp error flag indicating that the subject has neglected to inhale the dose measured by the metering assembly. The external device 15 has a display and software for visually displaying usage parameters through a GUI on the display. The usage parameters are stored as personalized data stored for predicting the future risk of exacerbation based on real-time data.

[0122] Figure 3A is a flowchart of a method 20 according to an embodiment of the present invention. The method 20 is executed by a system such as the system 10 shown in FIG. 1 and / or FIG. 2. For example, one or more of the first and / or second inhalers and / or the external device 15 and / or the personal data storage device 17 are configured to execute all or part of the method 20. That is, any combination of steps 22, 24, and 26 is executed by any combination of the first inhaler and / or the second inhaler and / or the external device 15 and / or the personal data storage device 17. Further, it should be understood that steps 22 and 24 are executed in any chronological order.

[0123] The method 20 has a step 22 of measuring the number of emergency inhalations of an emergency drug performed by a subject within a first period. In step 24, parameters regarding the airflow during at least several times, for example, during each emergency and / or regular inhalation, are measured. In step 26, a weighted model is used to calculate the probability of COPD exacerbation based on the number of emergency inhalations and the parameters. The model is weighted such that the parameters are more significant than the number of emergency inhalations in the probability calculation.

[0124] Although not described in the method 20, the system 10 is configured to notify the user when the probability of COPD exacerbation exceeds or is lower than a threshold. For example, the system 10 is configured to determine whether the probability reaches or exceeds a predetermined upper limit and / or whether the probability reaches or is lower than a predetermined lower limit. Accordingly, the system 10 is configured to treat the patient, for example, by switching the patient's treatment plan to a treatment plan configured for a higher (lower) risk of COPD exacerbation than the original treatment plan (e.g., through a message to the healthcare provider in charge of the patient).

[0125] System 10 notifies the user of the probability of exacerbation of COPD through one or more techniques. For example, System 10 is configured to display a message on the display of external device 15, send a message to a healthcare provider or third party related to the user, and activate an indicator (e.g., a light or speaker) of inhaler 100 to notify the user.

[0126] In the non-limiting example shown in FIG. 13A, the method further has step 23 of receiving an input of an indicator regarding the condition of the respiratory disease suffered by the subject. This input is used to increase the accuracy of the exacerbation prediction, as already explained.

[0127] In one embodiment, method 20 has a step of outputting a prompt to the user to prompt the user's input. The prompt is output based on the number of inhalations and inhalation parameters, but not based on the indicator. For example, the prompt is output based on an initial probability that has reached or exceeded a predetermined threshold. In this way, the user is prompted by the system to enter an indicator based on the initial probability indicating that an imminent exacerbation may occur. And when the user enters an indicator, the (overall) probability considering the indicator helps to confirm or justify the initial probability.

[0128] FIG. 3B is a flowchart and timeline regarding a method according to another embodiment of the present invention. The timeline shows the predicted day of occurrence of exacerbation ("day [0]"), 5 days before the exacerbation ("day [-5]"), and 10 days before the exacerbation ("day [-10]").

[0129] In FIG. 3B, block 222 represents an inhaler use notification regarded as a notification regarding the use of rescue medicine and / or maintenance medicine. Block 224 represents a notification of the flow rate corresponding to the parameter regarding the airflow during inhalation. Block 225 represents a "use" and "flow rate" notification regarded as a combined notification based on the use of the inhaler and inhalation parameters.

[0130] Block 226 represents a prediction notification. This prediction notification is based on the calculation of the initial probability described above. FIG. 3B shows that a questionnaire was presented in block 223 on day [-10]. The presentation of the questionnaire includes outputting a prompt that encourages the user to enter an indicator through the questionnaire. Block 227 represents indicating that the risk of deterioration remains even after the user input, based on the results of the questionnaire. This means that in block 230, the presentation of the questionnaire is continued, or the user is requested to enter the indicator again, or further input regarding the condition of their respiratory disease is requested. Block 231 represents that the risk of deterioration remains, for example, even after the calculation of the overall probability described above, and block 233 represents a scenario where the notification of the prediction of deterioration continues.

[0131] Block 228 represents a scenario where, after the presentation of the questionnaire in block 223, the calculated risk of deterioration returns to the baseline based on the indicators input by the user. Correspondingly, the risk notification ends in block 229.

[0132] Similarly, block 232 represents a scenario where, after the continued or further presentation of the questionnaire in block 230 ends, the risk of deterioration returns to the baseline. (For the sake of simplicity of explanation) Although not shown in FIG. 3B, the risk notification ends in block 232 after the risk of deterioration has returned to the baseline.

[0133] More generally, method 20 further has a step of preparing a first inhaler for administering a first aid drug to the subject. The first inhaler has a usage detection system configured to detect an inhalation performed by the subject using the first inhaler.

[0134] The number of emergency inhalations and parameters are measured by a usage detection system and a sensor system included in a first inhaler for administering an emergency drug, respectively. The sensor system selectively or additionally measures parameters regarding the airflow during regular inhalation of a maintenance drug using a second inhaler, as already explained.

[0135] Here, the weighted model supporting the method according to one embodiment of the present invention was the result of a clinical trial. This will be explained next. The following examples should be considered non-limiting examples for illustration.

[0136] The results of the study were generally applicable to another emergency drug administered using a different form of device, but albuterol administered using the ProAir Digihaler (registered trademark) commercially available from Teva Pharmaceutical Industries was used in a 12-week multi-site open-label trial.

[0137] The Digihaler was capable of recording the total number of inhalations, maximum inhalation flow rate, time to maximum inhalation flow rate, inhalation volume, and inhalation duration. Data was downloaded from the electronic module of the Digihaler at the end of the trial.

[0138] An acute exacerbation of COPD (AECOPD) was the primary outcome measure for this trial. In this trial, an AECOPD consisted of the onset of either a "severe AECOPD" or a "moderate AECOPD". In this trial, a "mild AECOPD" was not used as an AECOPD outcome measure.

[0139] A severe AECOPD is defined as an event that includes treatment with systemic corticosteroids (SCS, at least 10 mg of prednisone equivalent or more above the baseline value) and / or systemic antibiotics for at least two consecutive days, and a worsening of a respiratory disease that requires hospitalization for the treatment of AECOPD.

[0140] Concurrent AECOPD is defined as an event that includes treatment with SCS (at least 10 mg of prednisone equivalent or more above the reference value) and / or systemic antibiotics for at least 2 consecutive days and requires unscheduled contact for the treatment of AECOPD (such as phone calls, hospital visits, urgent care visits or emergency department visits), but does not require hospitalization, for the exacerbation of respiratory diseases.

[0141] Patients (aged 40 years or older) with COPD were recruited for the study. Patients used ProAir Digihaler as needed (90 mcg of albuterol as sulfate with lactose carrier, inhaled 1 - 2 times every 4 hours).

[0142] The following selection criteria were required. That is, the patient had developed at least one of concurrent or severe AECOPD in the last 12 months prior to screening, had received at least one administration of LABA, ICS / LABA, LAMA or LABA / LAMA in addition to SABA, and could demonstrate appropriate use of albuterol from the inhaler, and could discontinue the use of all other rescue or maintenance SABA or short - acting antimuscarinic agents during the study period and switch them to the inhaler provided for the study.

[0143] Patients were excluded from the study if, in the opinion of the investigator, they had clinically significant symptoms (treated or untreated) that interfered with participation in the study, for example, if they had underlying lung disorders that caused other confusions other than COPD, if they had used the investigational drug within either the 5 half - lives after dosing interruption or the longer of the two periods of 1 month with 2 visits, if they had developed congestive heart failure, if they were pregnant or lactating, or if there was a possibility of pregnancy during the study.

[0144] Two sub - populations consisting of approximately 100 patients were asked to wear an accelerometer on the ankle to measure physical activity (Total Daily Steps; TDS) or on the wrist to measure sleep disorder (Sleep Disturbance Index; SDI).

[0145] Interesting general factors regarding rescue medications were (1) The total number of inhalations per day immediately preceding the peak of AECOPD (2) The number of days when the number of albuterol uses increased immediately preceding the peak of AECOPD (3) The number of albuterol uses in 24 hours immediately preceding AECOPD as follows.

[0146] Approximately 400 patients were enrolled, resulting in 366 valuable patients who completed the trial. 336 effective inhalations with the DigiHaler were recorded. Further details on this are shown in Table 1.

[0147]

Table 1

[0148] 98 patients who completed the trial experienced AECOPD events and used the DigiHaler. A total of 121 sequential / severe AECOPD events were recorded. Further details are shown in Table 2.

[0149]

Table 2

[0150] For the 366 patients who completed the trial, 30 (8%) patients did not use an inhaler at all, 268 (73%) patients inhaled up to 5 times a day on average, and 11 (3%) patients inhaled more than 10 times a day on average.

[0151] Figure 4 is a graph 30a of the average number of rescue inhalations per subject versus the number of days from an exacerbation of COPD. Figure 4 shows data within a risk period of 14 days before and after the day of exacerbation. The straight line 32a corresponds to the average number of rescue inhalations per day within the risk period. The straight line 32a is at a higher position on the Y-axis than the average number of rescue inhalations per day outside the reference risk period, represented by the straight line 34a. This represents that the average number of rescue inhalations per day increases as the risk of exacerbation increases. For reference, Figure 4 further shows the average number of rescue inhalations per day of patients who did not experience the reference exacerbation, represented by the straight line 36a.

[0152] Figure 5 is another graph 30a of the average number of rescue inhalations per subject versus the number of days from an exacerbation of COPD. Figure 5 shows data within a risk period of 30 days before and after the day of exacerbation. Figure 5 shows that as the day of exacerbation approaches, the number of uses of the rescue inhaler increases significantly compared to the average number of rescue inhalations per day outside the reference risk period, represented by the straight line 34a.

[0153] The data indicates that the number of inhalations of rescue medications increases during the approximately two weeks prior to the occurrence of an exacerbation. The rate of increase is small during the approximately one week prior to the occurrence of an exacerbation. Table 3 shows further details regarding the relationship between the increased number of uses of rescue medications and AECOPD.

[0154]

Table 3

[0155] [1] For patients who experienced an AECOPD event, albuterol use was made before the symptoms of the event reached their maximum. For patients who experienced multiple events, only the first event was included in the analysis. The reference albuterol use is defined as the average number of inhalations during the first 7 days of the study. If no inhalation was performed during these first 7 days, the first available inhalation after these 7 days was used. If no inhalation was performed at all during the study period, the reference value is zero. [2]All of inferential statistics, odds ratio, p-value, and C statistic for goodness-of-fit are calculated as test variables using the increasing number of albuterol uses and the reference number of albuterol uses from the logistic regression model. An odds ratio greater than 1 indicates that patients whose daily albuterol use exceeds the reference value by more than 20% are more likely to experience AECOPD events than patients whose daily albuterol use does not exceed the reference value by more than 20%. Patients who experienced AECOPD between Day 1 and Day 7 of the trial were excluded from the analysis.

[0156] Figure 6 is a graph 40a of the average maximum inspiratory flow per subject against the number of days from the exacerbation of COPD. Figure 6 shows data for a risk period of 14 days before and after the day of exacerbation. The straight line 42a is slightly higher on the Y-axis than the average maximum inspiratory flow outside the reference risk period, represented by the straight line 44a, but this difference does not seem very important. Figure 6 further shows the average maximum inspiratory flow for patients who did not experience the reference exacerbation, represented by the straight line 46a.

[0157] Figure 7 is another graph 40a of the average maximum inspiratory flow per subject against the number of days from the exacerbation of COPD. Figure 7 shows data for 30 days before and after the day of exacerbation. Figure 7 shows a relatively stable, low average maximum inspiratory flow before the exacerbation.

[0158] Figure 8 is a graph 60a of the average inspiratory volume per subject against the number of days from the exacerbation of COPD. Figure 8 shows data within a 14-day risk period before and after the day of exacerbation. The straight line 62a corresponds to the average inspiratory volume within the risk period. The straight line 62a is at a lower position on the Y-axis than the average inspiratory volume within the reference risk period, represented by the straight line 64a. Figure 8 further shows the average inspiratory volume for subjects who did not experience the reference exacerbation, represented by the straight line 66a.

[0159] FIG. 9 is another graph 60a of the average inhalation volume per subject versus the number of days from the exacerbation of COPD. FIG. 9 shows data for a period of 30 days before and after the day of exacerbation.

[0160] FIG. 10 is a graph of the average inhalation duration per subject versus the number of days from the exacerbation of COPD. FIG. 10 shows data within a risk period of 14 days before and after the day of exacerbation. The straight line 72a corresponds to the average inhalation duration within the risk period. The straight line 72a is located lower on the Y-axis than the average inhalation duration outside the reference risk period, represented by the straight line 74a. FIG. 10 further shows the average inhalation duration for subjects who did not experience the reference exacerbation, represented by the straight line 76a.

[0161] FIG. 11 is another graph 70a of the average inhalation duration per subject versus the number of days from the exacerbation of COPD. FIG. 11 shows data within a period of 30 days before and after the day of exacerbation.

[0162] FIGS. 8 - 11 reveal a relatively long-term (demonstrated over about 30 days) linear decrease in the inhalation volume and inhalation duration before AECOPD.

[0163] Table 4 compares the inhalation parameters and the number of inhalations of rescue medications recorded for patients inside and outside the ±14-day AECOPD window with those recorded for patients who did not experience AECOPD.

[0164]

Table 4

[0165] The reference average number of albuterol inhalations per day for patients who experienced exacerbation was slightly higher compared to those who did not experience exacerbation, and the reference average inhalation volume and inhalation duration for patients who experienced exacerbation were slightly lower compared to those who did not experience exacerbation. Within the ±14-day AECOPD window, patients had a higher average number of albuterol inhalations per day compared to the reference value (outside the ±14-day AECOPD window) and also compared to patients who did not experience AECOPD.

[0166] It was found that the strongest predictors of COPD exacerbation are parameters related to airflow, such as peak inspiratory flow and / or inhalation volume and / or inhalation duration. Also, the number of rescue inhalations was found to contain significant predictors.

[0167] Based on the above results, a weighted model was developed to calculate the probability of COPD exacerbation. A supervised machine learning technique, Gradient Boosting Trees, was used to solve the classification problem (whether an exacerbation will occur in the next x days (exacerbation period)).

[0168] Gradient boosting techniques are well known in the prior art. See J.H. Friedman, Computational Statistics & Data Analysis, Vol. 38(4), pp. 367 - 378, 2002, and J.H. Friedman et al., The Annals of Statistics, Vol. 28(2), pp. 337 - 407, 2000. It produces a prediction model in the form of an ensemble (a plurality of learning algorithms) of basic prediction models consisting of decision trees (tree-shaped models and their possible conclusions). It forms a single powerful learner model in an iterative manner by using an optimization algorithm to minimize some appropriate loss function (a function of the difference between the estimated value and the true value for an example of data). The optimization algorithm calculates the expected value of the loss function using a training set of known values of the response variable (whether deterioration will occur in the next x days) and their corresponding predicted values (a list of characteristics and designed characteristics). The learning process successively fits new models and provides more accurate predicted values of the response variable.

[0169] Table 5 shows an exemplary list of factors included in the weighted model, along with their relative weightings with respect to each other.

[0170]

Table 5

[0171] The generalized model was evaluated by receiver operating characteristic (ROC) curve analysis. The most significant factor in the prediction model for calculating the probability of an impending COPD exacerbation is the inhalation parameter, but the prediction model became more certain by supplementing this with data on the number of rescue inhalations. Figure 12 shows the receiver operating characteristic (ROC) curve analysis of the model, evaluating the quality of the model by plotting the true positive rate against the false positive rate. This model used a value of 0.77 for the area under the ROC curve (AUC) to predict impending exacerbation over the next 5 days.

[0172] The number of rescue inhalations is a significant factor in improving the accuracy of the exacerbation probability calculation, even though it has a smaller impact on the overall probability than the inhalation parameters.

[0173] More generally, the first period over which the number of rescue inhalations should be measured is 1 to 30 days, preferably 5 to 15 days. Monitoring the number of rescue inhalations over the first period is particularly effective in calculating the probability of exacerbation of COPD.

[0174] When the inhalation parameter includes the maximum inspiratory flow rate, method 20 further has a step of calculating a maximum inspiratory flow rate, such as a minimum or average maximum inspiratory flow rate, from the maximum inspiratory flow rate measured for inhalations performed within a second period. The term "second" regarding the second period is intended to distinguish the period of sampling of the maximum inspiratory flow rate from the first period over which the number of rescue inhalations is sampled. The second period can at least partially overlap with the first period, or the first and second periods may coincide.

[0175] Thus, step 26 of calculating the probability of exacerbation of COPD is partially based on the minimum or average maximum inspiratory flow rate. The second period is selected according to the time required to collect appropriate reference value maximum inspiratory flow rate data in a manner similar to the considerations described above for the first period.

[0176] The calculation of the probability of exacerbation of COPD is partially based on, for example, as shown in FIGS. 6 and 7, the change with respect to the maximum inspiratory flow rate serving as the reference for the minimum or average maximum inspiratory flow rate.

[0177] For an improved accuracy in predicting exacerbation, the change in the maximum inspiratory flow rate serving as the reference for the minimum or average maximum inspiratory flow rate is, for example, 10% or more (such as 50% or more or 90% or more, etc.). The reference value is calculated using the minimum maximum inspiratory flow rate for one day measured over a period during which no exacerbation occurred, for example, 1 to 20 days, preferably 10 days. Optionally or additionally, the minimum or average maximum inspiratory flow rate is evaluated with respect to the absolute value.

[0178] Method 20 has a step of calculating an inspiratory volume such as a minimum or average inspiratory volume from the inspiratory volume measured for an inspiration performed within a third period. The term "third" with respect to the third period is intended to distinguish the period for sampling the inspiratory volume from a first period in which the number of rescue inhalations is sampled and a second period in which the maximum inspiratory flow rate data is sampled. The third period may at least partially overlap with the first period and / or the second period, or the third period may coincide with at least one of the first period and the second period.

[0179] Thus, step 26 of calculating the probability of exacerbation of COPD is partially based on the minimum or average inspiratory volume. The third period is selected according to the time required to collect the minimum inspiratory volume data of an appropriate reference value in a manner similar to the considerations described above for the first period.

[0180] The calculation of the probability of exacerbation of COPD is partially based on the change in the inspiratory volume serving as the reference for the minimum or average inspiratory volume, as shown, for example, in FIGS. 8 and 9.

[0181] For an improved accuracy in predicting exacerbation, the change in the inspiratory volume serving as the reference for the minimum or average inspiratory volume is, for example, 10% or more (such as 50% or more or 90% or more, etc.). The reference value is calculated using the minimum inspiratory volume for one day measured over a period during which no exacerbation occurred, for example, 1 to 20 days, preferably 10 days. Optionally or additionally, the minimum or average maximum inspiratory flow rate is evaluated with respect to the absolute value.

[0182] Method 20 has a step of calculating a minimum or average inhalation duration from the inhalation durations measured for inhalation over a fourth period. The term "fourth" with respect to the fourth period is intended to distinguish the period for sampling the inhalation duration from a first period in which the number of rescue inhalations is sampled, a second period in which the maximum inhalation flow rate data is sampled, and a third period in which the inhalation volume data is sampled. The fourth period can at least partially overlap with the first period and / or the second period and / or the third period, or the fourth period may coincide with at least one of the first period, the second period, and the third period.

[0183] Thus, step 26 of calculating the probability of exacerbation of COPD is partially based on the minimum or average inhalation duration. The fourth period is selected according to the time required to collect minimum inhalation duration data of appropriate reference values in a manner similar to the considerations described above for the first period.

[0184] The calculation of the probability of exacerbation of COPD is partially based on, for example, as shown in FIGS. 10 and 11, the change in the inhalation duration serving as a reference for the minimum or average inhalation duration.

[0185] For the improved accuracy of predicting exacerbation, the change in the inhalation duration serving as a reference for the minimum or average inhalation duration is, for example, 10% or more (such as 50% or more or 90% or more, etc.). The reference value is calculated using the minimum inhalation duration of one day measured over a period during which no exacerbation occurred, for example, 1 to 20 days, preferably 10 days. Optionally or additionally, the minimum or average maximum inhalation duration is evaluated with respect to the absolute value.

[0186] Another clinical trial was conducted to better understand the factors affecting the prediction of asthma exacerbation. The examples described below should be regarded as non-limiting examples for one explanation.

[0187] Although the results of the test are generally applicable to other emergency medications administered using another form of device, albuterol administered using ProAir Digihaler® (registered trademark), marketed by Teva Pharmaceutical Industries, was used in an open-label trial over 12 weeks.

[0188] Patients (18 years of age or older) with asthma prone to exacerbation were recruited for the test. The patients used ProAir Digihaler as needed (inhaled 90 mcg of albuterol as sulfate with lactose carrier 1 - 2 times every 4 hours).

[0189] The electronic module of the digihaler recorded parameters regarding each use, i.e., each inhalation and the airflow between each inhalation, i.e., maximum inhalation flow rate, inhalation volume, time to maximum inhalation flow rate, and inhalation duration. The data was downloaded from the inhaler and applied to a machine learning algorithm for developing a model to predict impending exacerbation, together with clinical data.

[0190] The diagnosis of clinical asthma exacerbation (CAE) in this example was based on the American Thoracic Society / European Respiratory Society statement (H.K. Reddel et al., Am J Respir Crit Care Med. 2009, Vol. 180(1), pp. 59 - 99). It included both "severe CAE" and "moderate CAE".

[0191] Severe CAE was defined as CAE that included exacerbation of asthma requiring administration of oral steroids (prednisone or its equivalent) for at least 3 days and hospitalization. Conventional CAE requires at least 3 days of dosing or hospitalization with oral steroids (prednisone or its equivalent).

[0192] The generalized model was evaluated by receiver operating characteristic (ROC) curve analysis, as will be described in more detail below with reference to FIG. 17.

[0193] The objectives of the study and the primary endpoint were to examine, alone or in combination with other study data such as parameters regarding airflow, physical activity, and sleep during inhalation preceding CAE, the usage pattern and amount of albuterol preceding CAE captured by the DigiHaler. This study provides the first successful attempt to develop a model for the prediction of CAE derived from the use of an emergency inhaler equipped with an integrated sensor capable of measuring inhalation parameters.

[0194] FIG. 13 shows three timelines showing different inhalation patterns recorded by their respective DigiHalers for three different patients. The top timeline shows that the patient in question makes one inhalation at a time. The bottom timeline shows that the patient in question makes two or more consecutive inhalations in one session. Here, the term "session" is defined as a series of inhalations where consecutive inhalations are made at time intervals within 60 seconds. The middle timeline shows that the patient in question makes inhalations in various patterns. That is, the DigiHaler is configured to record not only the number of emergency inhalations but also the usage pattern.

[0195] It was found that 360 patients had made one or more valid inhalations from the DigiHaler. These 360 patients were included in the analysis. Of these patients, 64 patients experienced a total of 73 cases of CAE. Figure 14 is a graph 30 of the average number of emergency inhalations against the number of days since the exacerbation of asthma. Figure 14 shows data for a 14-day risk period before and after the day of exacerbation. The straight line 32 corresponds to the average number of emergency inhalations per day during the risk period. The straight line 32 is located higher on the Y-axis than the average number of emergency inhalations per day outside the reference risk period, represented by the straight line 34. This indicates that the average number of emergency inhalations per day increases as the risk of exacerbation increases. For reference, Figure 14 also shows the average number of emergency inhalations per day for patients who did not experience the reference exacerbation, represented by the straight line 36.

[0196] Figure 15 is another graph 30 of the average number of emergency inhalations against the number of days since the exacerbation of asthma. Figure 15 shows data within a 50-day period before and after the day of exacerbation. Figure 15 shows that the number of times the inhaler is used increases significantly as the day of exacerbation approaches, compared to the average number of emergency inhalations per day outside the reference risk period, represented by the straight line 34.

[0197] Figure 16 is four graphs of the rate of change of various parameters related to the number of emergency inhalations and airflow against the number of days since the exacerbation of asthma, plotted against the reference values of each parameter.

[0198] Graph 40 plots the rate of change of the number of emergency inhalations (outside the risk period) against the number of days since the exacerbation of asthma, relative to the reference value. It was found that the number of emergency inhalations increased by 90% relative to the reference value immediately before the exacerbation.

[0199] Graph 42 plots the rate of change of the reference value of the minimum peak inspiratory flow per day against the number of days since the exacerbation of asthma. Graph 42 shows that the minimum peak inspiratory flow per day generally decreases day by day until the exacerbation occurs. It was found that the minimum peak inspiratory flow per day decreased by 12% relative to the reference value immediately before the exacerbation.

[0200] Graph 44 plots the number of days from exacerbation against the rate of change with respect to the reference value of the minimum daily inhalation volume. Graph 44 shows that the minimum daily inhalation volume generally decreases day by day until exacerbation. It was found that the minimum daily inhalation volume decreases by 20% with respect to the reference value immediately before exacerbation.

[0201] Graph 46 plots the number of days from exacerbation against the rate of change with respect to the reference value of the minimum daily inhalation duration. Graph 46 shows that the minimum daily inhalation duration generally decreases day by day until exacerbation. It was found that the minimum daily inhalation duration decreases by 15% - 20% with respect to the reference value immediately before exacerbation.

[0202] When constructing the first weighted prediction model, it was found that, particularly within the 5 - day period prior to CAE, the strongest predictor for exacerbation of asthma is the average number of rescue inhalations per day. It was also found that the parameters related to airflow, namely, the maximum inhalation flow rate and / or inhalation volume and / or inhalation duration, also have significant predictive values.

[0203] In the first weighted model, it was found that the most significant characteristics in calculating the probability of exacerbation of asthma are that the number of rescue inhalations is 61%, the inhalation tendency is 16%, the maximum inhalation flow rate is 13%, the inhalation volume is 8%, and the use of albuterol at night is 2%. Such inhalation characteristics are collected by the DigiHaler, and the maximum inhalation flow rate, time - to - maximum inhalation flow rate, inhalation volume, inhalation duration, night - time use, and the trends of these parameters over time are recorded.

[0204] The inhalation tendency is an artificially created or designed parameter, such as the rate of change of today's inhalation volume compared to the last three days. Another example is the rate of change of the number of rescue inhalations today compared to the last three days. In these examples, each tendency is the change in the inhalation volume and the number of rescue inhalations respectively, rather than the inhalation volume and the number of rescue inhalations themselves.

[0205] Based on the above results, a first weighted prediction model was developed to calculate the probability of exacerbation of asthma. Supervised machine learning techniques and gradient boosting were used to solve the classification problem (whether exacerbation occurs in the next x days (exacerbation period)). The gradient boosting technique was used in the same manner as described above for the COPD exacerbation prediction model.

[0206] Table 6 exemplarily lists the factors included in the first weighted prediction model, along with their relative weights to each other.

[0207]

Table 6

[0208] Figure 17 is a diagram showing the receiver operating characteristic (ROC) curve analysis of the model, which evaluates the quality of the model by plotting the true positive rate against the false positive rate. This first weighted prediction model predicted imminent exacerbation in the subsequent 5 days with an AUC value of 0.75 using the relevant characteristics described above. When characteristics based only on the number of rescue inhalations were used, the AUC value was 0.69.

[0209] A second weighted prediction model was developed using the same data in an effort to improve the first weighted prediction model for predicting exacerbation of asthma. Parameters related to the living body were included in the modeling. In particular, case report form (CRF) data such as medical history, body mass index (BMI), and blood pressure were combined with the data of Dihalal and applied to a machine learning algorithm to improve the accuracy of the prediction model.

[0210] The algorithm was trained not only with respect to age, BMI, blood pressure, and the number of exacerbations and hospitalizations in the most recent 12 months, but also with respect to patient-specific inhalation information collected by the digital haler. The comparison between the reference characteristics and the characteristics before prediction, and the tendency of the change of those characteristics were applied to the supervised machine learning algorithm. The four-fold cross-validation technique was used to compare the performance measurement criteria, and gradient boosting was selected as the optimal algorithm. As before, the generalized model was evaluated by the analysis of the receiver operating characteristic curve (ROC) and the area under the curve (AUC).

[0211] Table 7 exemplarily lists the factors included in the second weighted prediction model, together with their relative weighting to each other.

[0212]

Table 7

[0213] The second weighted prediction model predicted the impending exacerbation in the subsequent 5 days with an AUC value of 0.83. The second weighted prediction model has a sensitivity of 68.8% and a specificity of 89.1%. Thus, it was shown that the second weighted prediction model is an improved asthma exacerbation prediction model compared to the first weighted prediction model with an AUC value of 0.75 described above. The additional improvement of the second weighted prediction model is at least partially attributed to the introduction of parameters related to the living body.

[0214] An important factor in the prediction model for calculating the probability of impending exacerbation of asthma is the number of rescue inhalations and the tendency regarding the number of rescue inhalations, but it was found that the prediction model becomes more certain by supplementing this factor with parameters related to the airflow during inhalation.

[0215] Therefore, although the influence on probability is smaller than the number of emergency inhalations, the parameters related to the airflow during inhalation are shown to be significant factors that increase the calculation accuracy of the probability of asthma exacerbation, similar to other factors other than the number of emergency inhalations.

[0216] Figures 18 to 22 show non-limiting examples of inhalers included in the system 10.

[0217] Figure 18 is a perspective view of a first inhaler 100 according to a non-limiting example. The inhaler 100 is, for example, a self-inhaled inhaler. The inhaler 100 has an upper cap 102, a main housing 104 and / or a mouthpiece 106 and / or a mouthpiece cover 108 and / or an electronic module 120 and / or a ventilation port 126. The mouthpiece cover 108 is coupled to the main housing 104 by a hinge, and the mouthpiece 106 is exposed or housed by opening and closing the mouthpiece cover. In this example, a hinge coupling is used, but the mouthpiece cover 108 can also be coupled to the inhaler 100 by another type of coupling. Further, in this example, the electronic module 120 is housed in the upper cap 102 at the upper part of the main housing 104, but the electronic module 120 may be integrated with the main housing 104 of the inhaler 100 and / or housed inside the main housing 104.

[0218] Figure 19 is a longitudinal sectional view of the inhaler 100 shown in Figure 18. The inhaler 100 has, inside the main housing 104, a drug tank 110 (for example, a hopper), bellows 112, a bellows spring 114, a yoke (not shown), a dosing cup 116, a dosing chamber 117, a crusher 121 and a flow path 119. The drug tank 110 contains a drug such as a dry powder drug for administration to a subject. When the mouthpiece cover 108 is moved from the closed position to the open position, the bellows 112 supplies a single dose of the drug from the drug tank to the dosing cup 116. Thereafter, the subject inhales through the mouthpiece 106 and aspirates a single dose of the drug.

[0219] Due to the airflow generated by the inhalation of the subject, the crusher 121 aerosolizes the single-dose medicine by crushing the aggregates of the medicine in the dosing cup 116. The crusher 121 is configured to aerosolize the medicine when the airflow through the flow path 119 reaches or exceeds a specific speed, or when it has a speed within a specific range. When aerosolized, the single-dose medicine flows from the dosing cup 116 through the flow path 119 into the dosing chamber 117, and then is supplied to the subject from the mouthpiece 106. When the airflow through the flow path 119 does not reach or exceed a specific speed, or when the speed is not within a specific range, the medicine remains in the dosing cup 116. If the medicine in the dosing cup 116 is not aerosolized by the crusher 121, when the mouthpiece cover 108 is subsequently opened, the next single-dose of the medicine is not supplied from the medicine tank 110. That is, the single-dose of the medicine remains in the dosing cup until it is aerosolized by the crusher 121. When a single-dose of the medicine is administered, the confirmation of the administration is stored in the memory as dosing confirmation information in the inhaler 100.

[0220] When the subject inhales through the mouthpiece 106, air enters from the ventilation port, generating an airflow for supplying the medicine to the subject. The flow path 119 extends from the dosing chamber 117 to the end of the mouthpiece 106 and includes the inner parts of the dosing chamber 117 and the mouthpiece 106. The dosing cup 116 is inside or adjacent to the dosing chamber 117. Further, the inhaler 100 has a dosing counter 111 configured such that the total number of doses of the medicine in the medicine tank 110 is initially set and is decremented by one each time the mouthpiece cover 108 is moved from the closed position to the open position.

[0221] The upper cap 102 is attached to the main housing 104. The upper cap 102 is attached to the main housing 104 using, for example, one or more clips that can engage with a recess in the main housing 104. The upper cap 102, when joined, overlaps a part of the main housing 104 and, for example, a substantial airtight seal is created between the upper cap 102 and the main housing 104.

[0222] FIG. 20 is an exploded perspective view of the inhaler 100 shown in FIG. 18, showing a state in which the upper cap 102 is removed and the electronic module 120 is exposed. As shown in FIG. 20, the upper surface of the main housing 104 has one or more (e.g., two) orifices 146. One orifice 146 is configured to receive the slider 140. For example, when the upper cap 102 is attached to the main housing 104, the slider 140 projects from the upper surface of the main housing 104 through one orifice 146.

[0223] FIG. 21 is an exploded perspective view of the upper cap 102 and the electronic module of the inhaler 100 shown in FIG. 18. As shown in FIG. 21, the slider 140 forms an arm 142, a stopper 144, and a distal end 145. The distal end 145 forms the bottom of the slider 140. The distal end 145 of the slider 140 is configured to abut against a yoke inside the main housing 104 (e.g., when the mouthpiece cover 108 is in the closed position or a partially open position). The distal end 145 is configured to abut against the upper surface of the yoke when the yoke is in either of the radial directions. For example, the upper surface of the yoke has a plurality of apertures (not shown), and the distal end 145 of the slider 140 is configured to abut against the upper surface of the yoke regardless of whether one aperture is aligned with the slider 140 in a row.

[0224] The upper cap 102 has a slider guide 148 configured to receive a slider spring 146 and a slider 140. The slider spring 146 is located inside the slider guide. The slider spring 146 engages the inner surface of the upper cap 102 and engages the upper portion (e.g., proximal end) of the slider 140. When the slider 140 is disposed inside the slider guide 148, the slider spring 146 is partially compressed between the upper portion of the slider 140 and the inner surface of the upper cap 102. For example, the slider spring 146 is configured such that when the mouthpiece cover 108 is closed, the distal end 145 of the slider 140 contacts the yoke. The distal end 145 of the slider 145 is also configured to contact the yoke while the mouthpiece cover 108 is being opened or closed. The stopper 144 of the slider 140 engages the slider guide 148 such that, for example, the slider 140 is retained within the slider guide 148 as the mouthpiece cover 108 is opened and closed or vice versa. The stopper 144 and the slider guide 148 are configured to limit the vertical (e.g., axial) movement of the slider 140. This limited movement distance is shorter than the limited movement distance of the vertical movement of the yoke. That is, when the mouthpiece cover 108 is moved to the fully open position, the yoke continues to move vertically towards the mouthpiece 106, but the stopper 144 stops the vertical movement of the slider 140, whereby the distal end 145 of the slider 140 no longer contacts the yoke.

[0225] More generally, the yoke is mechanically coupled to the mouthpiece cover 108 and compresses the bellows spring when the mouthpiece cover 108 is opened from the closed position, and then releases the compressed bellows spring 114 when the mouthpiece cover 108 reaches the open position, thereby causing a single dose of the drug to be supplied from the drug tank 110 to the dosing cup 116 in the bellows 112. The yoke is in contact with the slider 140 when the mouthpiece cover 108 is in the closed position. The slider 140 is moved by the yoke when the mouthpiece cover 108 is opened from the closed position and is provided to be separated from the yoke when the mouthpiece cover 108 reaches the fully open position. This configuration is considered a non-limiting example of the metering assembly already described above. This is because when the mouthpiece cover 108 is opened, a single dose of the drug is metered.

[0226] By the movement of the slider 140 during the metering of the drug, the slider 140 engages the switch 130 and activates the switch 130. The switch 130 activates the electronic module 120 and records the metering of the drug. The slider 140 and the switch 130, together with the electronic module 120, correspond to a non-limiting example of the usage detection system 12B described above. In this example, the slider 140 is considered a means configured such that the usage detection system 12B records a single dose of the drug by the metering assembly, and each metering thereby represents an inhalation performed by the subject using the inhaler 100.

[0227] By the activation of the switch 130 by the slider 140, for example, the electronic module 120 transitions from the first power state to the second power state and detects an inhalation from the mouthpiece 108 by the subject.

[0228] The electronic module 120 has a printed circuit board (PCB) assembly 122 and / or a switch 130 and / or a power source (e.g., a battery 126) and / or a battery holder 124. The PCB assembly 122 has a sensor system 128 and / or a wireless communication circuit 129 and / or a switch 130 and / or surface-mounted components such as one or two or more light-emitting diodes (LEDs) (not shown), such as indicators. The electronic module 120 has a controller (e.g., a processor) and / or a memory. The controller and / or the memory are physically different from the components of the PCB 122. Alternatively, the controller and the memory consist of a part of another chipset mounted on the PCB 122. For example, the wireless communication circuit 129 includes a controller and / or a memory for the electronic module 120. The controller of the electronic module 120 includes a microcontroller, a programmable logic device (PLD), a microprocessor, an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or any suitable processing device or control circuit.

[0229] The controller reads data from the memory and records data in the memory. The memory includes any suitable form of memory, such as non-removable memory and / or removable memory. Non-removable memory includes random access memory (RAM), read-only memory (ROM), a hard disk, or any other form of storage device. Removable memory includes subscriber identity module (SIM) cards, memory sticks, secure digital (SD) memory cards, and the like. The memory is disposed inside the controller. The controller also reads data from a memory that is not physically disposed inside the electronic module 120, such as on a server or a smartphone, and records data therein.

[0230] The sensor system 128 has one or more sensors. The sensor system 128 is an example of the sensor system 12A. The sensor system 128 has one or more different types of sensors, such as, for example, one or more pressure sensors and / or temperature sensors and / or humidity sensors and / or direction sensors and / or acoustic sensors and / or optical sensors. The one or more pressure sensors include barometric pressure sensors (e.g., atmospheric pressure sensors) and / or differential pressure sensors and / or absolute pressure sensors and / or the like. The sensors use microelectromechanical system (MEMS) and / or nanoelectromechanical system (NEMS) technologies. The sensor system 128 is configured to provide instantaneous measurement values (e.g., pressure measurement values) and / or measurement values aggregated over a certain period of time (e.g., pressure measurement values) to the controller of the electronic module 120. As shown in FIGS. 19 and 20, the sensor system 128 is disposed outside the flow path 119 of the inhaler 100 but is pneumatically coupled to the flow path.

[0231] The controller of the electronic module 120 receives a signal corresponding to the measurement value from the sensor system 128. The controller calculates or determines metrics regarding the air flow using the signal received from the sensor system 128. The metrics regarding the air flow represent the profile of the air flow through the flow path 119 of the inhaler 100. For example, when the sensor system 128 records an air pressure change of 0.3 kilopascals (kPa), the electronic module 120 calculates that the air pressure change corresponds to an air flow rate of approximately 45 liters per minute (Lpm) through the flow path 119.

[0232] FIG. 22 is a graph of air flow rate versus pressure. The air flow rate and profile shown in FIG. 22 are merely exemplary, and the calculated speeds depend on the size, shape, and design of the inhaler 100 and its components.

[0233] The controller of the electronic module 120 generates personalized data in real time as part of an evaluation of how the inhaler 100 was used and / or whether its use is likely to result in complete administration of the drug by comparing the signals received from the sensor system 128 and / or the calculated metrics regarding the airflow to one or more thresholds or numerical ranges. For example, when the calculated metric regarding the airflow corresponds to an inhalation with an air flow rate lower than a specific threshold, the electronic module 120 determines that there is no inhalation from the mouthpiece 106 of the inhaler 100 or that the inhalation from the mouthpiece 106 is insufficient. Also, when the calculated metric regarding the airflow corresponds to an inhalation with an air flow rate exceeding a specific threshold, the electronic module 120 determines that the inhalation from the mouthpiece 106 of the inhaler 100 is excessive. Further, when the calculated metric regarding the airflow corresponds to an inhalation with an air flow rate within a specific range, the electronic module 120 determines that the inhalation is good or that there is a possibility that complete dosing has resulted.

[0234] The pressure measurement value and / or the calculated metric regarding the airflow represents the quality or strength of the inhalation from the inhaler 100. For example, when compared to a specific threshold or numerical range, the measurement value and / or the metric are used to classify the relevant inhalation into certain event types such as a good inhalation event, a low inhalation event, a no inhalation event, or an excessive inhalation event. The classification of the inhalation is a useful parameter recorded as personalized data for the subject.

[0235] An inhalation event is associated with metrics of pressure measurements and / or airflow below a specific threshold, such as an air flow rate less than 30 Lpm. An inhalation event occurs when the subject does not inhale from the mouthpiece 106 after opening the mouthpiece cover 108 and during the measurement cycle. Also, an inhalation event occurs when the inhalation by the subject generates only an insufficient airflow to operate the nebulizer 121, i.e., to aerosolize the drug in the dosing cup 116, and is insufficient to ensure proper administration of the drug through the flow path 119.

[0236] A low inhalation event is associated with metrics of pressure measurements and / or airflow within a specific numerical range, such as 30 Lpm to 45 Lpm. A low inhalation event occurs when the subject inhales from the mouthpiece 106 after opening the mouthpiece cover 108, and only at least a portion of the drug for one dose is administered through the flow path 119 by the subject's inhalation. That is, the inhalation is insufficient to operate the nebulizer 121 to aerosolize at least a portion of the drug from the dosing cup 116.

[0237] A good inhalation event is associated with metrics of pressure measurements and / or airflow above that of a low inhalation event, such as 45 Lpm to 200 Lpm. A good inhalation event occurs when the subject inhales from the mouthpiece 106 after opening the mouthpiece cover 108, and a sufficient inhalation is made by the subject's effort to ensure proper administration of the drug through the flow path 119, such as when a sufficient airflow is generated by the subject's effort to operate the nebulizer 121 to aerosolize all of the drug for one dose in the dosing cup 116.

[0238] An over-inhalation event is associated with metrics related to pressure measurements and / or airflow above those of a good inhalation event, such as an airflow rate exceeding 200 Lpm. An over-inhalation event occurs when the subject inhales beyond the normal operating parameters of the inhaler 100. An over-inhalation event can also occur even if the subject inhales within the normal range, when the inhaler 100 is not placed or held in the appropriate position during use. For example, the calculated airflow rate exceeds 200 Lpm when the ventilation port is blocked or obstructed (e.g., by one finger or the thumb) while the subject is inhaling from the mouthpiece 106.

[0239] Any suitable threshold or numerical range is used to classify specific events. Some events or all events may be used. For example, a no-inhalation event is associated with an airflow rate lower than 45 Lpm, and a good inhalation event is associated with an airflow rate within the range of 45 Lpm to 200 Lpm. Thus, a low-inhalation event may not be used in some cases.

[0240] The metrics related to the pressure measurement and / or the calculated airflow also represent the direction of the airflow through the flow path 119 of the inhaler 100. For example, when the pressure measurement reflects a negative pressure change, the pressure measurement represents an airflow exiting the mouthpiece 106 through the flow path 119. When the pressure measurement reflects a positive pressure change, the pressure measurement represents an airflow entering the mouthpiece 106 through the flow path 119. Therefore, the metrics related to the pressure measurement and / or the airflow are used to determine whether the subject is exhaling air into the mouthpiece 106, thereby indicating that the subject is not using the inhaler 100 correctly.

[0241] Inhaler 100 has a spirometer or a device that performs a similar operation to enable measurement of metrics related to lung function. For example, inhaler 100 makes measurements to obtain metrics related to the vital capacity of a subject. The spirometer or a device that performs a similar operation measures the volume of air inhaled and / or exhaled by the subject. The spirometer or a device that performs a similar operation detects changes in the volume of air inhaled and / or exhaled using a pressure transducer, ultrasound, or a water level gauge.

[0242] Personalized data (e.g., metrics related to pressure, airflow, lung function, dosing confirmation information, etc.) collected from the use of inhaler 100 or calculated based on the use of inhaler 100 is further calculated and / or evaluated (partially or wholly) by an external device. More generally, the wireless communication circuit 129 of the electronic module 120 has not only a transmitter and / or a receiver (e.g., a transceiver), but also additional electrical circuits. For example, the wireless communication circuit 129 has a Bluetooth chipset (e.g., a Bluetooth low power chipset), a ZigBee chipset, a Thread chipset, and so on. Thus, the electronic module 120 wirelessly provides personalized data such as pressure measurement values and / or metrics related to airflow and / or metrics related to lung function and / or dosing confirmation information and / or other circumstances related to the use of the inhaler 100 to an external device including a smartphone. The personalized data is provided to the external device in real time, thereby enabling prediction of the exacerbation risk based on personalized data related to the user of the inhaler, such as real-time data indicating the number of uses from the inhaler 100 and how the inhaler 100 was used, and real-time data related to the lung function and / or treatment of the subject. The external device has software for processing the received information and providing compliance feedback to the user of the inhaler 100 via a graphical user interface (GUI).

[0243] Metrics related to airflow include personalized data collected in real time from inhaler 100, such as the average flow rate of inhalation / exhalation and / or the maximum flow rate of inhalation / exhalation (e.g., the maximum inhalation flow rate received) and / or the inhalation / exhalation volume and / or the time to inhalation / exhalation maximum and / or the duration of inhalation / exhalation, among one or more of these. Metrics related to airflow also represent the direction of the airflow through flow path 119. That is, a negative pressure change corresponds to inhalation from the mouthpiece 106, while a positive pressure change corresponds to exhalation into the mouthpiece 106. In calculating the metrics related to airflow, the electronic module 120 is configured to remove or minimize the distortion caused by environmental conditions. For example, the electronic module 120 performs calculations to account for changes in atmospheric pressure before or after calculating the metrics related to airflow. One or more pressure measurements and / or metrics related to airflow are recorded in the memory of the electronic module 120 after being time-stamped.

[0244] In addition to the metrics related to airflow, the inhaler 100 or another computing device generates additional personalized data using the metrics related to airflow. For example, the controller of the electronic module 120 of the inhaler 100 translates the metrics related to airflow into other metrics representing the lung function and / or lung health of the subject, which are understood to be medical parameters such as the maximum inspiratory flow rate and / or the maximum expiratory flow rate and / or the forced expiratory volume in one second (FEV1). The electronic module 120 of the inhaler 100 calculates a measure of the subject's lung function and / or lung health using a mathematical model such as a regression model. The mathematical model identifies the correlation between the total inhalation volume and FEV1. The mathematical model identifies the correlation between the maximum inhalation flow rate and FEV1. The mathematical model identifies the correlation between the total inhalation volume and the maximum expiratory flow rate. The mathematical model identifies the correlation between the maximum inhalation flow rate and the maximum expiratory flow rate.

[0245] The battery 126 supplies power to the components of the PCB 122. The battery 126 consists of any suitable power source for supplying power to the electronic module 120, such as a coin cell, for example. The battery 126 can be rechargeable or non-rechargeable. The battery 126 is housed within a battery holder 124. The battery holder 124 is attached to the PCB 122 such that the battery 126 is in continuous contact with the PCB 122 and / or in electrical contact with the components of the PCB 122. The battery 126 has a specific battery capacity that affects the lifespan of the battery 126. As will be further described below, the distribution of power from the battery 126 to one or more components of the PCB 122 is managed such that the battery 126 is guaranteed to supply power to the electronic module 120 over the service life of the inhaler 100 and / or the shelf life of the drug contained therein.

[0246] With the connection established, the communication circuit and the memory are powered on, and the electronic module 120 is paired with an external device such as a smartphone. The controller retrieves data from the memory and wirelessly transmits that data to the external device. The controller retrieves and transmits the most recently stored data in the memory. The controller also retrieves and transmits a portion of the most recently stored data in the memory. For example, the controller can determine which portions of the data have already been transmitted to the external device and then transmit the portions of the data that have not yet been transmitted. Alternatively, the external device requests specific data from the controller, such as data collected by the electronic module 120 after a certain period of time or after the most recent transmission to the external device. The controller, if any, retrieves that specific data from the memory and transmits the retrieved data to the external device.

[0247] Data stored in the memory of the electronic module 120 (e.g., signals generated by switches and / or pressure measurements obtained by the sensor system 128 and / or metrics related to air flow calculated by the controller of the PCB 122) is transmitted to an external device. The external device processes and analyzes the data and calculates useful parameters related to the inhaler 100. Further, a mobile application on a mobile device generates feedback for the user based on the data received from the electronic module 120. For example, the mobile application generates daily, monthly, and annual reports and / or confirms or notifies the subject of error events and / or provides useful feedback to the subject and / or does similar things.

[0248] The embodiments described above are merely examples for explaining the configuration of the present invention and thus do not limit the configuration of the present invention. Other variations of the embodiments described above can be understood and provided by those skilled in the art from studying the drawings, the specification, and the appended claims in the implementation of the present invention. The fact that specific means are defined in different cited form claims does not mean that combinations of those means cannot be used advantageously.

Claims

1. 1. A system for calculating the probability of a COPD exacerbation in a subject, comprising: a first inhaler for administering a rescue medication to the subject, the first inhaler having a usage detection system configured to measure rescue inhalations performed by the subject using the first inhaler; and an optional second inhaler for administering a maintenance medication to the subject between scheduled inhalations; a sensor system configured to measure a parameter related to airflow during the rescue inhalation and / or the regular inhalation using the second inhaler when the second inhaler is used in the system; and a processor configured to measure a number of the rescue inhalations within a first time period, to receive measurements of the parameter during at least some of the rescue and / or regular inhalations, and to calculate, using a weighted model, a probability of an exacerbation of the COPD based on the measurements of the parameter during the rescue inhalations; The model is weighted such that measurements of the parameters are more significant in calculating the probability than the number of rescue inhalations.

2. 2. The system of claim 1, wherein the probability of a COPD exacerbation is the probability that the exacerbation will occur within an exacerbation period following the first period.

3. 3. The system according to claim 1, wherein the first period is from 1 to 30 days.

4. The system according to any one of claims 1 to 3, wherein the parameter is at least one of a maximum inhalation flow rate, an inhalation volume, and an inhalation duration.

5. 5. The system of claim 4, wherein the processor is further configured to calculate an average peak flow rate from the peak flow rates measured during the rescue inhalation and / or the scheduled inhalation within a second period of time, and wherein the probability of a COPD exacerbation is calculated based in part on the average peak flow rate, and optionally the second period of time is between 1 and 30 days.

6. 6. The system of claim 5, wherein the processor is configured to calculate the probability of a COPD exacerbation based in part on a change in average peak inspiratory flow relative to a baseline peak inspiratory flow.

7. The system of any one of claims 4 to 6, characterized in that the processor is further configured to calculate an average inhalation volume from the inhalation volumes measured during the emergency inhalations and / or regular inhalations performed within a third period of time, optionally the third period of time being between 1 and 30 days.

8. 8. The system of claim 7, wherein the processor is configured to calculate the probability of a COPD exacerbation based in part on a change in the average inhalation volume relative to a baseline inhalation volume.

9. The system of any one of claims 4 to 8, wherein the processor is configured to calculate an average inhalation duration from inhalation durations measured during rescue inhalation and / or regular inhalation within a fourth period of time, and the probability of a COPD exacerbation is calculated in part based on the average inhalation duration, optionally wherein the fourth period of time is between 1 and 30 days.

10. 10. The system of claim 9, wherein the processor is configured to calculate the probability of a COPD exacerbation based in part on a change in the average inhalation duration relative to a baseline inhalation duration.

11. A system as described in any one of claims 1 to 10, characterized in that the sensor system comprises a pressure sensor, and optionally the usage detection system comprises another pressure sensor, the pressure sensor and the another pressure sensor having the same or different configurations from each other.

12. The first inhaler comprises: A chemical tank, a metering assembly configured to meter a dose of drug from the drug reservoir; The system of any one of claims 1 to 11, characterized in that the usage detection system is configured to record the metering of the dose of medication by the metering assembly, whereby the metering of the dose of medication is indicative of the rescue inhalation performed by the subject's use of the first inhaler.

13. A user interface for receiving an input of an index representing a state of a respiratory disease suffered by the subject, The system according to any one of claims 1 to 12, characterized in that the processor is configured to calculate the probability of an exacerbation of the COPD using the weighted model based on the number of rescue inhalations and the measured values ​​of the parameters and the indicator that has been input.

14. 1. A method for calculating the probability of a COPD exacerbation in a subject, comprising: receiving a number of rescue inhalations of rescue medication performed by the subject within a first time period; receiving measurements of a parameter related to airflow during at least some of the rescue inhalations and / or regular inhalations of a maintenance medication performed by the subject; determining a probability of a COPD exacerbation based on the number of rescue inhalations and the measured parameters using a weighted model; The method of claim 1, wherein the model is weighted such that measurements of the parameters are more significant than the number of rescue inhalations in calculating the probability.

15. 15. The method of claim 14, further comprising providing a first inhaler for administering the rescue medication to the subject, the first inhaler having a usage detection system configured to identify the rescue inhalation being performed by the subject using the first inhaler.

16. 16. The method of claim 14 or claim 15, further comprising providing a sensor system configured to measure said parameters relating to airflow during said rescue inhalation and / or said regular inhalation.

17. 15. A computer program comprising a computer program code adapted to perform the method according to claim 14, when the computer program runs on a computer.

18. 1. A method of treating COPD in a subject, comprising: Carrying out the method according to any one of claims 14 to 16, determining whether the probability reaches or exceeds a predetermined upper limit, or determining whether the probability reaches or is below a predetermined lower limit; treating the COPD based on the probability of reaching or exceeding the upper limit or based on the probability of reaching or falling below the lower limit.

19. 20. The method of claim 18, wherein the treatment comprises switching the subject from a first treatment plan to a second treatment plan based on the probability that the upper limit has been reached or exceeded, the second treatment plan being designed to address a higher risk of COPD exacerbations than the first treatment plan.

20. 20. The method of claim 19, wherein the second treatment regimen includes administration of a biopharmaceutical, optionally comprising any one of omalizumab, mepolizumab, reslizumab, benralizumab, and dupilumab, or a combination of two or more thereof.

21. 20. The method of claim 18, wherein the treatment comprises switching the subject from a first treatment plan to a third treatment plan based on the probability reaching or being below the lower limit, the third treatment plan being designed to address a lower risk of COPD exacerbations than the first treatment plan.

22. 1. A method for diagnosing an exacerbation of COPD, comprising: Carrying out the method according to any one of claims 14 to 16, determining whether the probability reaches or exceeds a predetermined upper limit indicative of an exacerbation of the COPD; and diagnosing an exacerbation of said COPD based on said probability of said upper limit being reached or exceeded.

23. 1. A method for delimiting a subpopulation of a subject, comprising: Carrying out the method according to any one of claims 14 to 16 for each of the subjects belonging to the population of subjects; thereby calculating said probability for each said subject belonging to said population; setting a probability threshold or range of probabilities that distinguishes the probability calculated for the subpopulation from the probability calculated for the remainder of the population; defining said subpopulation from the remainder of said population using said probability threshold or said probability range.

Images