Method and system for referring a Patient
A computer-implemented referral method using clinical guidelines automates the referral process, improving clinician efficiency, reducing waiting times, and ensuring appropriate referrals, thus enhancing patient safety and resource utilization.
Patent Information
- Application Number
- GB2023018345
- Authority / Receiving Office
- GB · GB
- Patent Type
- Applications
- Current Assignee / Owner
- Filing Date
- 2023-11-30
- Publication Date
- 2025-09-10
Abstract
Description
Technical Field
[0001] The present disclosure relates to a referral process. In particular, the present invention relates to a computer-implemented method and a computing apparatus for the referral of a patient. Background
[0002] A primary healthcare professional such as a general practitioner (GP), dentist, optician, or pharmacist is typically the first point of contact for a person with a health problem. A GP practice, for example, receives requests for help or advice through various means such as a patient phone conversation, in person visit, or through an online consultation. However, a GP practice may not be equipped with the relevant resources and / or specialist knowledge to deal with the large volume of requests with varying patient needs, and health backgrounds. Thus, for each patient request the GP practice needs to determine: why the help was sought from the GP, what kind of help the patient needs, how quickly the patient needs help, and where and when the patient should be seen. This initial assessment is an essential step that allows a GP practice to efficiently deal with the large volume of requests, allowing the GP practice to direct the patient to the right type of care based on the patient’s needs. For example, the GP may refer the patient to the relevant department of a hospital or a clinic which is equipped with the relevant resources to address the patient’s needs and / or the GP may refer the patient to a specialist who is equipped with the relevant specialist knowledge to address the patient’s needs. In some instances, the GP may inadvertently refer the patient to a department or specialist which may not be best equipped to deal with the patient’s condition. This in turn leads to a delay in treatment for the patient, where any delay in receiving treatment may lead to discomfort for the patient and may be detrimental to the patient’s health and safety.
[0003] The above-mentioned referral may be made through an electronic referral system such as the NHS e-Referral Service, in the UK, (e-RS) by the primary healthcare professional to the relevant department and / or specialist. The referral may be in the form of a letter, and may detail clinical information corresponding to the patient. The referral may also detail the relevant information relating to the specialist, diagnostic imaging service, or pathology service such as a blood test, as recommended by the primary healthcare professional to address the patient’s needs. The referral is then manually assessed and triaged by, for example, a consultant or clinician from the department receiving the referral in order to determine the appropriate care pathway for the patient, or if the referral is deemed not to be appropriate, then the consultant or clinician may decide to send the referral back to the primary healthcare professional with appropriate feedback. The assessment by the consultant or clinician from the receiving department is important as it ensures that the received referral from the primary healthcare professional is best suited to address the patient’s needs or condition which in turn ensures clinical safety. The assessment by the consultant or clinician also ensures appropriate resource allocation of the department receiving the referral, as referrals which are not deemed to be optimal to address the patient’s needs or condition are returned to the primary healthcare professional with appropriate feedback.
[0004] However, the current practice of triaging referrals as described above has several drawbacks. As mentioned above, the consultant or clinician from the receiving department must manually assess the referral. This results in the consultant or clinician spending significant amount of time reviewing a large volume of referrals which in turn leads to the consultant or clinician not being able to provide direct patient care. The manual nature of the current method also leads to slower access of appointments for patients. Typically, the average time taken to triage a referral may be two days from submission, which may be detrimental to a patient’s health and safety. Delays may be especially detrimental for patients with major health conditions such as cancer, where any delay in treatment can increase the risk of death. Furthermore, as will be appreciated, consultants or clinicians work in a high-pressured environment where they are having to manually review a large set of referrals. This may in some cases lead to the consultant or clinician accepting inappropriate referrals instead of returning the referral back to the referrer with appropriate feedback. Not only does this lead to unnecessary diagnostics and use of resources but may also be detrimental to a patient’s safety, as the consultant or clinician may have overlooked important clinical information relating to the patient. This may put the patient at risk of further exacerbating their condition as the referral deemed to be approved by the consultant or clinician may not be one that is suitable to address the patient’s needs.
[0005] Accordingly, the current method of triaging referrals consumes a significant amount of consultant or clinician time, provides slow access to appointments, and leads to unnecessary diagnostics and appointments, all of which may be detrimental not only to the patient’s health but also to the patient’s safety.
[0006] Furthermore, the current practice leads to increased variability in triaging decisions between different organisations and clinicians / consultants due to, for example, a variation in local practices between the different organisations and clinicians / consultants. This results in variable access to services, including attendances of low clinical value. This in turn leads to a delay in the patient receiving the necessary treatment to address their needs or condition where, as explained above, any delay in treatment may be detrimental to a patient’s health and safety.
[0007] The present application seeks to mitigate or overcome at least one or more of the problems described above. Summary
[0008] According to an aspect of the invention a computer-implemented method is disclosed. The method is suitable for referring a patient. The method comprises receiving a referral from a referrer, the referral comprising referral text and a referrer input. The method further comprises extracting clinical information from the referral text. The method further comprises analysing at least the extracted clinical information from the referral text. The method further comprises generating a referral output as a function of at least the analysed extracted clinical information. The method further comprises comparing the referral output to the referrer input. The method further comprises determining if the referral output corresponds to the referrer input, wherein if the referral output corresponds to the referrer input, the referrer input is approved, wherein if the referrer input is approved, the method further comprises generating an output as a function of the correspondence of the referral output to the referrer input, wherein the generated output comprises the approved referrer input and outputting the generated output.
[0009] Advantageously, the above method for referring a patient frees up consultant or clinician time for direct patient care as the consultant or clinician does not have to manually review a large volume of referral letters from the primary healthcare professional. This results in effective use of the consultant’s or clinician’s time leading to improved patient care and safety. In addition, the referral is also triaged quicker leading to reduced waiting times for patients which in turn also leads to improved patient care and safety. Furthermore, the above method avoids the approval of inappropriate referrals as there is no requirement for a consultant or clinician to manually review a large volume of referral letters, which as explained above leads to unnecessary diagnostics and appointments and may be detrimental to a patient’s safety. Thus, the automatic nature of the above method leads to an efficient use of resources and increased safety for patients.
[0010] Accordingly, the method for referring a patient as described herein frees up a significant amount of consultant or clinician time, provides faster access to appointments, and avoids unnecessary diagnostics and appointments, all of which lead to improved patient health and safety.
[0011] The method may further comprise determining if the referral output corresponds to the referrer input, wherein if the referral output does not correspond to the referrer input, the referrer input is not approved, wherein if the referrer input is not approved, the method may further comprise manually reviewing and triaging the referral, where the manual review and triage may be conducted by a clinician or a consultant, sending the referral back to the referrer with appropriate feedback, or directing the referral to the department or service as determined by the generated referral output. Embodiments are described in relation to Figure 1 below.
[0012] The term “referrer” as used herein is understood to mean a health practitioner such as a GP, dentist, optician or a pharmacist and may also be an allied health practitioner such as a nurse practitioner, midwife, physiotherapist, osteopath, optometrist or psychologist.
[0013] The term “referral” as used herein is understood to mean a request made by any of the above mentioned healthcare practitioners to a service, department and / or specialist equipped with the relevant resources and knowledge to further assess a patient’s condition and / or provide the necessary care and treatment. Thus, the term “GP referral” is understood to mean a request made by a GP to a service, department and / or specialist equipped with the relevant resources and knowledge to further assess a patient’s condition and / or provide the necessary care and treatment. The referral may be made using an electronic referral system such as the NHS e-RS. The referral may be in the form of a letter. The referral may comprise referral text. The referral may comprise a referrer input. The referral may comprise advice and guidance corresponding to treatment of the patient’s condition. The skilled person would appreciate that the referral may be made using any electronic or non-electronic means, where the non-electronic means may be a letter sent via post.
[0014] Receiving a referral from a referrer may comprise receiving a pending referral from a primary healthcare professional via, for example, the NHS e-RS. The referrer may be a GP and the referral may be a GP referral. The receiving may be performed, for example, by one or more processors of a computing apparatus (such as that shown in Figure 3). The skilled person would appreciate that the referral may also be obtained from any other electronic referral system.
[0015] The terms “consultant” or “clinician” as used herein is understood to mean a consultant or clinician from the department receiving the referral who reviews the referral in order to determine the appropriate care pathway or sends the referral back to the primary healthcare professional with appropriate feedback if the referral is not appropriate to address the patients’ needs or condition. The terms “consultant” and “clinicians” are used interchangeably.
[0016] The term “referral text” as used herein is understood to mean a body of text within the referral letter. The referral text may comprise patient information corresponding to a patient. The patient information may comprise at least one of: patient’s name, age, gender, symptoms and medical history.
[0017] The term “referrer input” as used herein is understood to mean a service referral and / or an urgency level as determined by a primary healthcare professional such as a GP. The referrer input may comprise at least one of a service referral and / or an urgency level as determined by a primary healthcare professional such as a GP.
[0018] The term “referral output” as used herein is understood to mean a service level and / or an urgency level as determined by, for example, one or more processors of a computing apparatus (such as that shown in Figure 3). The referral output may comprise at least one of a service level and / or an urgency level as determined by, for example, one or more processors of a computing apparatus (such as that shown in Figure 3). The referral output may be determined using clinical guidelines. The clinical guidelines may be from a regulatory body, a nationally recognised body or an independent organisation.
[0019] The service referral may comprise a recommendation of a service such as a diagnostic imaging service, or pathology service such as a blood test or treatment. The service referral may comprise recommendation of diagnosis by a specialist clinician as determined by a primary healthcare professional such as a GP or by, for example, one or more processors of a computing apparatus (such as that shown in Figure 3). The service referral may reflect the action required in order to address the patient’s needs. The urgency level may be understood to mean the time frame in which the specified action in the service referral must be performed as determined by a primary healthcare professional such as a GP or by, for example, one or more processors of a computing apparatus (such as that shown in Figure 3).
[0020] Generating an output as a function of the correspondence of the referral output to the referrer input may comprise accessing a library of local services and determining a relevant local service as a function of the approved referrer input. This is explained further in relation to Figure 1 below. This advantageously allows for an efficient means to determine the services available at the hospital receiving the referral.
[0021] Extracting clinical information from the referral text may comprise using a language processing technique to extract clinical information from the referral text. The language processing technique may be natural language processing (NLP). The skilled person would appreciate that other language processing techniques may also be used to extract clinical information from the referral text. This advantageously allows for a fast, and efficient means to extract key information from a referral.
[0022] The term “clinical information” as used herein may be understood to mean any symptom(s) associated with the patient that may be extracted from the referral text as defined above. Clinical information may also relate to any other information relating to the patient which is extracted from the referral text which allows for the assessment of the patient’s condition and in turn for the generation of a referral output by a computing apparatus such as the computing apparatus of Figure 3. In an example, only clinical information may be used to generate a referral output. In another example, patient information and clinical information may be used in combination to generate a referral output.
[0023] Analysing at least the extracted clinical information from the referral text may comprise comparing the extracted clinical information with a database. The database may comprise information relating to clinical guidelines. The clinical guidelines may be from a regulatory body, a nationally recognised body or an independent organisation. The regulatory body may be NICE. Analysing at least the extracted clinical information may be performed, for example, by one or more processors of a computing apparatus (such as that shown in Figure 3). This is explained further in relation to Figure 1 below.
[0024] The database may be a repository comprising information relating to clinical guidelines. The clinical guidelines may be from a regulatory body, a nationally recognised body or an independent organisation.
[0025] The inventors have recognised that clinical guidelines are a useful tool which can be utilised to obtain a referral destination for a patient. Clinical guidelines are systematically developed by, for example, a regulatory body such as the National Institute for Health and Care Excellence (NICE) to assist health practitioners to make decisions about appropriate healthcare pathways for specific clinical circumstances. Clinical guidelines offer concise instructions relating to, for example, which diagnostic or screening test to order, how to provide medical or surgical services, how long a patient should stay in the hospital, or other details relating to clinical practice. That is, the guidelines provide recommendations on how healthcare and other professionals should care for people with specific conditions, where the recommendations are based on the best available evidence. The clinical guidelines may cover any aspect of a condition. This may include recommendations about: providing information, education and advice (for example, about self-care), prevention, treatment in primary care (GPs and other community services), treatment in secondary care (provided by or in hospitals), treatment in specialised services.
[0026] As such, the inventors have recognised that clinical guidelines may be utilised to obtain a referral destination which can then be compared to, for example, the service, test or specialist as recommended in the referral from the primary healthcare professional in order to automate the process relating to triaging of a referral. This is explained in further detail in relation to Figure 1 below.
[0027] Furthermore, triaging based on clinical guidelines such as the NICE guidelines and clinical best practices helps reduce practice variation between different organisations and clinicians / consultants. Triaging based on clinical guidelines also leads an output which is based on the best available evidence in relation to the patient’s condition. As such, triaging based on clinical guidelines provides a reliable and optimal referral output to address the patient’s needs, which in turn leads to improved patient health and safety.
[0028] According to an aspect of the invention, a computer-readable medium is described herein. The computer-readable medium has instructions stored thereon which, when executed by one or more processors, cause the one or more processors to implement a method for referring a patient as described above. The computer-readable medium may comprise a non-transitory computer-readable medium. The computer-readable medium may comprise, for example, a USB stick, a hard drive, or some other memory unit.
[0029] According to an aspect of the invention, a computing apparatus is provided herein. The computing apparatus is suitable for referring a patient. The apparatus comprises one or more memory units. The computing apparatus further comprises one or more processors configured to execute instructions stored in the one or more memory units to perform a method for referring a patient as described above.
[0030] According to an an aspect of the invention a computer-implemented method is disclosed. The method is suitable for referring a patient. The method comprises receiving a referral from a referrer, the referral comprising referral text. The method further comprises extracting clinical information from the referral text. The method further comprises analysing at least the extracted clinical information from the referral text. The method further comprises generating a referral output as a function of at least the analysed extracted clinical information. The method further comprises accepting or rejecting the referral based on the generated referral output.
[0031] Advantageously, the above method for referring a patient frees up consultant or clinician time for direct patient care as the consultant or clinician does not have to manually review a large volume of referral letters from the primary healthcare professional. This results in effective use of the consultant’s or clinician’s time leading to improved patient care and safety. In addition, the referral is also triaged quicker leading to reduced waiting times for patients which in turn also leads to improved patient care and safety. Furthermore, the above method avoids the approval of inappropriate referrals as there is no requirement for a consultant or clinician to manually review a large volume of referral letters, which as explained above leads to unnecessary diagnostics and appointments and may be detrimental to a patient’s safety. Thus, the automatic nature of the above method leads to an efficient use of resources and increased safety for patients.
[0032] Accordingly, the method for referring a patient as described herein frees up a significant amount of consultant or clinician time, provides faster access to appointments, and avoids unnecessary diagnostics and appointments, all of which lead to improved patient health and safety.
[0033] The service referral may comprise a recommendation of a service such as a diagnostic imaging service, or pathology service such as a blood test or treatment and / or recommendation of diagnosis by a specialist clinician as determined by a primary healthcare professional such as a GP or by, for example, one or more processors of a computing apparatus (such as that shown in Figure 3). As explained above, the service referral may reflect the action required in order to address the patient’s needs. The urgency level may be understood to mean the time frame in which the specified action in the service referral must be performed as determined by a primary healthcare professional such as a GP or by, for example, one or more processors of a computing apparatus (such as that shown in Figure 3).
[0034] Receiving a referral from a referrer may comprise receiving a pending referral from a primary healthcare professional via, for example, the NHS e-RS. The referrer may be a GP and the referral may be a GP referral. The receiving may be performed, for example, by one or more processors of a computing apparatus (such as that shown in Figure 3). The skilled person would appreciate that the referral may also be obtained from any other electronic referral system.
[0035] Extracting clinical information from the referral text may comprise using a language processing technique to extract clinical information from the referral text. The language processing technique may be natural language processing (NLP). The skilled person would appreciate that other language processing techniques may also be used to extract clinical information from the referral text. This advantageously allows for a fast, and efficient means to extract key information from a referral.
[0036] Analysing at least the extracted clinical information from the referral text may comprise comparing the extracted clinical information with a database. The database may comprise clinical guidelines. The clinical guidelines may be from a regulatory body, a nationally recognised body or an independent organisation. The regulatory body may be NICE. Analysing at least the extracted clinical information may be performed, for example, by one or more processors of a computing apparatus (such as that shown in Figure 3).
[0037] The database may be a repository comprising information relating to clinical guidelines. The clinical guidelines may be from a regulatory body, a nationally recognised body or an independent organisation.
[0038] As explained above, triaging based on clinical guidelines such as the NICE guidelines and clinical best practices helps reduce practice variation between different organisations and clinicians / consultants. Triaging based on clinical guidelines also leads an output which is based on the best available evidence in relation to the patient’s condition. As such, triaging based on clinical guidelines provides a reliable and optimal referral output to address the patient’s needs, which in turn leads to improved patient health and safety.
[0039] Accepting the referral based on the generated referral output may comprise generating an output as a function of the referral output. Accepting the referral may further comprise outputting the generated output.
[0040] Generating an output as a function of the referral output may comprise accessing a library of local services and determining a relevant local service. This advantageously allows for an efficient means to determine the services available at the hospital receiving the referral.
[0041] Rejecting the referral based on the generated referral output may comprise sending the referral back to the referrer with appropriate feedback, manually reviewing and triaging the referral, where the manual review and triage may be conducted by a clinician or a consultant or redirecting the referral to the department or service as determined by the generated referral output.
[0042] According to an aspect of the invention, a computer-readable medium is disclosed. The computer-readable medium has instructions stored thereon which, when executed by one or more processors, cause the one or more processors to implement a method for referring a patient as described above. The computer-readable medium may comprise a non-transitory computer-readable medium. The computer-readable medium may comprise, for example, a USB stick, a hard drive, or some other memory unit.
[0043] According to an aspect of the invention, a computing apparatus is disclosed. The computing apparatus is suitable for referring a patient. The apparatus comprises one or more memory units. The computing apparatus further comprises one or more processors configured to execute instructions stored in the one or more memory units to perform a method for referring a patient as described above.
[0044] Many modifications and other embodiments of the present invention set out herein will come to mind to a person skilled in the art in light of the teachings presented herein. Therefore, it will be understood that the disclosure herein is not to be limited to the specific embodiments disclosed herein. Moreover, although the description provided herein provides example embodiments in the context of certain combinations of elements, steps and / or functions may be provided by alternative embodiments without departing from the scope of the invention.
[0045] The person skilled in the art will also appreciate that features described in relation to one aspect may be used in combination with another aspect, and vice versa. The person skilled in the art will also appreciate that features described in relation to one embodiment may be used in combination with another embodiment, and vice versa. Brief Description of the Figures
[0046] Embodiments of the invention will now be described by way of example only, with reference to the accompanying figures, in which: Figure 1 shows a flowchart of a method for referring a patient based on correspondence of the referral output to the referrer input; Figure 2 shows a flowchart of a method for referring a patient based on the referral output; and Figure 3 shows a block diagram of a computing apparatus.
[0047] Throughout the description and the drawings, like reference numerals refer to like parts. Detailed Description
[0048] The present disclosure provides methods for referring a patient and further discloses computing apparatuses configured to refer a patient. Whilst various embodiments are described below, the invention is not limited to these embodiments, and variations of these embodiments may fall within the scope of the invention which is to be limited only by the appended claims.
[0049] A computer-implemented method 100 for referring a patient will now be described in relation to the flowchart shown in Figure 1. The method may be performed by any suitable computing apparatus, such as the computing apparatus 300 described in relation to Figure 3 below.
[0050] At step 102, the method comprises receiving a referral from a referrer, the referral comprising referral text and a referrer input. In the embodiment of Figure 1, the referral is a GP referral. For example, a GP may use the e-RS system to make a referral for a patient. The GP referral is received by a computing apparatus 300, where the GP referral comprises information relating to the patient such as the patient’s name, age, gender, symptoms and / or medical history (referral text), information relating to the recommended service (service referral), and the urgency level as determined by the GP based on the GP’s diagnosis of the patient. In embodiments, the GP referral comprises at least one of patient information, urgency level and / or information relating to the recommended service.
[0051] In embodiments, the referral may be made using any electronic or non-electronic means, where the non-electronic means may be a letter sent via post. In embodiments, the referrer may be any health practitioner such as a GP, dentist, optician, or pharmacist and the patient information may comprise one or more of the patient’s name, age, gender, symptoms and / or medical history.
[0052] At step, 104 the method comprises extracting clinical information from the referral text. NLP is used to extract clinical information from the referral text, where the referral text comprises patient information such as the patient’s name, age, gender, symptoms and / or medical history. As an example, a referral from a GP may describe as part of the referral text that a 56-year-old patient suffers from “abdominal pain, dispepsia, and weight-loss... he denies any difficulty in swallowing food... patient contracted covid-19 a month ago”. NLP is used to extract terms such as “abdominal pain”, weight loss”, “covid-19” and in the above example, even correct terms like “dispepsia” to “dyspepsia”. The skilled person would appreciate that any other language processing technique which is common in the art may also be used to extract such terms.
[0053] At step 106, the method comprises analysing at least the extracted clinical information from the referral text. Analysing at least the extracted clinical information comprises comparing the extracted clinical information to a database in order to generate a referral output. In particular, the extracted clinical information is compared to guidelines from NICE to generate a referral output, as explained below. The skilled person would appreciate that the extracted information may be compared to clinical guidelines from other regulatory bodies, nationally recognised bodies or independent organisations to generate a referral output.
[0054] As highlighted above, the NICE guidelines are a set of clinical guidelines which assist health practitioners to make decisions about appropriate healthcare pathways for specific clinical circumstances. They offer concise instructions relating to, for example, which diagnostic or screening test to order, how to provide medical or surgical services, how long a patient should stay in the hospital, or other details relating to clinical practice. For example, the NICE guidelines in relation to oesophageal cancer define that direct access upper gastrointestinal endoscopy should be performed within 2 weeks to assess for oesophageal cancer in people with dysphagia or those aged 55 years and over with weight loss and any of the following: upper abdominal pain, reflux, or dyspepsia. Thus, if the extracted clinical information from the GP referral, in particular from the referral text, includes terms such as “abdominal pain”, weight loss”, and “dyspepsia” and / or patient information such as that the patient is 56 years old, then comparison with the guidelines can be used to determine a referral output. In this case, comparison with the NICE guidelines is used to determine a referral output which comprises endoscopy as a service referral and an urgency level relating to a 2-week period. As such, the one or more processors of a computing apparatus such as the computing apparatus of Figure 3 is configured to generate a referral output as a function of at least the analysed extracted clinical information (step 108). The skilled person would appreciate that in some cases only clinical information may be used to generate a referral output. In other examples, clinical information and patient information may be used to generate a referral output.
[0055] At step 108, the method comprises generating a referral output as a function of at least the analysed extracted clinical information.
[0056] At step 110, the method comprises comparing the referral output to the referrer input. That is, the referrer input in the GP referral is compared to the referral output as generated in step 108.
[0057] At step 112, determination is made as to whether or not the referral output corresponds to the referrer input. If yes, the referrer input is approved (step 114), and the method proceeds to steps 116 and 118. If no, the referrer input is not approved, and the method proceeds to step 120.
[0058] At step 116, the method comprises generating an output as function of the correspondence of the referral output to the referrer input, wherein the generated output comprises the approved referrer input. That is, if the referrer input which comprises the service referral and the urgency level as determined by the GP, corresponds to the referral output (service referral and urgency level) as generated by a computing apparatus in step 108, then an output is generated by the one or more processors of a computing apparatus such as the computing apparatus of Figure 3, where the generated output comprises the approved referrer input. This advantageously leads to the approval of a referral without the need for the manual review of a clinician which in turn leads to a referral process with improved safety for patients. The generated output is then output by the computing apparatus, as described below in step 118.
[0059] At step, 118, the method comprises outputting the generated output.
[0060] In embodiments, the output may be displayed on the visual display 330 of computing apparatus 300 and / or may be transmitted by the computing apparatus 300 to the relevant department as described below.
[0061] As such, the method as described herein advantageously frees up consultant or clinician time for direct patient care as the consultant or clinician does not have to manually review a large volume of referral letters from the primary healthcare professional. This results in effective use of the consultant’s or clinician’s time leading to improved patient care and safety. In addition, the referral is also triaged quicker leading to reduced waiting times for patients which in turn also leads to improved patient care and safety. Furthermore, the above method avoids the approval of inappropriate referrals as there is no requirement for a consultant or clinician to manually review a large volume of referral letters, which as explained above leads to unnecessary diagnostics and appointments and may be detrimental to a patient’s safety. Thus, the automatic nature of the above method leads to an efficient use of resources and increased safety for patients.
[0062] Accordingly, the method for referring a patient as described herein frees up a significant amount of consultant or clinician time, provides faster access to appointments, and avoids unnecessary diagnostics and appointments, all of which lead to improved patient health and safety.
[0063] In one example, if the GP referral comprises endoscopy as a service referral with a 2-week period as the urgency level and the referral output, as determined by a computing apparatus, also comprises endoscopy as a service referral with a2-week period, then the referrer input as determined by the GP is deemed to be an optimal referral and is considered an approved referrer input. The one or more processors of, for example, computing apparatus of Figure 3 then generate an output as a function of the correspondence of the referral output to the referrer input, wherein the generated output comprises the approved referrer input, which in the above example would be endoscopy as a service referral with a 2-week period as the urgency level. The one or more processors of, for example, computing apparatus of Figure 3 then output this generated output which may be sent to the relevant department equipped to deal with the referral. The relevant department may be determined by accessing a library of local services as described below.
[0064] In embodiments, generating an output as a function of the correspondence of the referral output to the referrer input may comprise accessing a library of local services and determining a relevant local service as a function of the approved referrer input. For example, once it is determined that the referrer input as determined by a GP is one which matches or corresponds to the referral output as determined using clinical guidelines, then a library of services provided by the hospital or clinic which received the referral (hereafter referred to as the receiving hospital or clinic) may be accessed by a computing apparatus. That is, once it is determined that the referrer input is approved, a library of local services provided by the hospital or clinic that received the referral may be accessed by a computing apparatus. The library of services provided by the receiving hospital or clinic is hereafter referred to as the library of local services. The library of local services may detail the departments, clinicians and / or services provided by the receiving hospital or clinic. The library of local services may detail the availability of appointments for the departments, clinicians and / or services provided by the local hospital or clinic. The library of local services may be a database stored on a server or the cloud comprising information relating to the departments, clinicians and / or services provided by the receiving hospital or clinic and the availability of appointments for the departments, clinicians and / or services provided by the local hospital or clinic. A computing apparatus such as the one shown in Figure 3 may access the library of local services and then determine the relevant local service. The computing apparatus may then generate an output as a correspondence of the referral output to the referrer input, wherein the generated output may comprise the approved referrer input and the determined local service.
[0065] For example, if the approved referrer input relates to endoscopy as a service referral to be performed within a 2-week period as the urgency level, then the computing apparatus such as the one shown in Figure 3 may access a library of local services relating to the receiving hospital or clinic and determine if the receiving hospital or clinic has the facilities or resources to perform the endoscopy referral within the 2-week period. Once it has been determined that the facilities or resources are available at the receiving hospital or clinic and the relevant local service has been selected, then the computing apparatus such as the one shown in Figure 3 may generate an output as a correspondence of the referral output to the referrer input, wherein the generated output comprises the approved referrer input and the determined local service. The computing apparatus may then output this generated output, where the output may be transmitted to the selected service within the receiving hospital or clinic by computing apparatus 300.
[0066] At step 120, when the referral output does not correspond to the referrer input, the referral output is manually assessed by a clinician in order to determine the relevant referral destination and the urgency level. That is, when the referral output does not correspond to the referrer input, the referrer input is manually triaged by the clinician. In another example, as a clinical safety net, complex and high-risk referrals may also be left for manual review.
[0067] In embodiments, when the referral output does not correspond to the referrer input, the referral may be sent back to the referrer with appropriate feedback or may be sent to the department or service as determined by the generated referral output. As an example, if the referrer input as determined by a GP indicates that the patient requires the services provided by the cardiology department i.e., it is a cardiology referral, however, the referral output as determined by a computing apparatus using clinical guidelines indicates that the patient should in fact be referred to a respiratory specialist then the referral is directed to the department and / or specialist equipped to deal with respiratory referrals.
[0068] The skilled person would appreciate that the above method is applicable to a referral made by any health practitioner such as a GP, dentist, optician, or pharmacist, where the referral may be made using any electronic or non-electronic means.
[0069] A computer-implemented method 200 for referring a patient will now be described in relation to the flowchart shown in Figure 2. The method may be performed by any suitable computing apparatus, such as the computing apparatus 300 described in relation to Figure 3 below. At step 210, the method comprises receiving a referral from a referrer, the referral comprising referral text and a referrer input. In the embodiment of Figure 2, the referral is a GP referral. For example, a GP may use the e-RS system to make a referral for a patient. The GP referral is received by a computing apparatus 300, where the GP referral comprises information relating to the patient such as the patient’s name, age, gender, symptoms and / or medical history (referral text), information relating to the recommended service (service referral) and the urgency level as determined by the GP based on the GP’s diagnosis of the patient. In embodiments, the GP referral comprises at least one of patient information, urgency level and / or information relating to the recommended service.
[0070] In embodiments, the referral may be made using any electronic or non-electronic means, where the non-electronic means may be a letter sent via post. In embodiments, the referrer may be any health practitioner such as a GP, dentist, optician, or pharmacist and the patient information may comprise one or more of the patient’s name, age, gender, symptoms and / or medical history.
[0071] At step, 220 the method comprises extracting clinical information from the referral text. NLP is used to extract clinical information from the referral text, where the referral text comprises patient information such as the patient’s name, age, gender, symptoms and / or medical history. As an example, a referral from a GP may describe as part of the referral text that a 56-year-old patient suffers from “abdominal pain, dispepsia, and weight-loss... he denies any difficulty in swallowing food... patient contracted covid-19 a month ago”. NLP is used to extract terms such as “abdominal pain”, weight loss”, “covid-19” from the GP referral and in the above example, even correct terms like “dispepsia” to “dyspepsia”. The skilled person would appreciate that any other language processing technique which is common in the art may also be used to extract such terms.
[0072] At step 230, the method comprises analysing at least the extracted clinical information from the referral text. Analysing at least the extracted clinical information comprises comparing the extracted clinical information to a database in order to generate a referral output. In particular, the extracted clinical information is compared to a database comprising information relating to the NICE guidelines to generate a referral output, as explained below. The skilled person would appreciate that the extracted information may be compared to clinical guidelines from other regulatory bodies, nationally recognised bodies or independent organisations to generate a referral output.
[0073] As explained above, the NICE guidelines are a set of clinical guidelines which assist health practitioners to make decisions about appropriate healthcare pathways for specific clinical circumstances. They offer concise instructions relating to, for example, which diagnostic or screening test to order, how to provide medical or surgical services, how long a patient should stay in the hospital, or other details relating to clinical practice. For example, the NICE guidelines in relation to oesophageal cancer define that direct access upper gastrointestinal endoscopy should be performed within 2 weeks to assess for oesophageal cancer in people with dysphagia or those aged 55 years and over with weight loss and any of the following: upper abdominal pain, reflux, or dyspepsia. Thus, if the extracted clinical information from the referral, in particular from the referral text, includes terms such as “abdominal pain”, weight loss”, and “dyspepsia” and / or patient information such as that the patient is 56 years old, then comparison with the guidelines can be used to determine a referral output. In this case, comparison with the NICE guidelines is used to determine a referral output which comprises endoscopy as a service referral and an urgency level relating to a 2-week period. As such, the one or more processors of a computing apparatus such as the computing apparatus of Figure 3 is configured to generate a referral output as a function of at least the analysed extracted clinical information (step 240). The skilled person would appreciate that in some cases only clinical information may be used to generate a referral output. In other examples, clinical information and patient information may be used to generate a referral output.
[0074] At step 240, the method comprises generating a referral output as a function of at least the analysed extracted clinical information.
[0075] At step 250, the method comprises accepting or rejecting the referral based on the generated referral output.
[0076] Accepting the referral comprises generating an output as a function of the referral output and outputting the generated output. This is explained below in further detail.
[0077] In embodiments, the generated output may be displayed on the visual display 330 of computing apparatus 300 and / or may be transmitted by the computing apparatus 300 to the relevant department as described below.
[0078] As such, the method as described herein advantageously frees up consultant or clinician time for direct patient care as the consultant or clinician does not have to manually review a large volume of referral letters from the primary healthcare professional. This results in effective use of the consultant’s or clinician’s time leading to improved patient care and safety. In addition, the referral is also triaged quicker leading to reduced waiting times for patients which in turn also leads to improved patient care and safety. Furthermore, the above method avoids the approval of inappropriate referrals as there is no requirement for a consultant or clinician to manually review a large volume of referral letters, which as explained above leads to unnecessary diagnostics and appointments and may be detrimental to a patient’s safety. Thus, the automatic nature of the above method leads to an efficient use of resources and increased safety for patients.
[0079] Accordingly, the method for referring a patient as described herein frees up a significant amount of consultant or clinician time, provides faster access to appointments, and avoids unnecessary diagnostics and appointments, all of which lead to improved patient health and safety.
[0080] In one example, if comparison with the NICE guidelines is used to determine a referral output which comprises endoscopy as a service referral and an urgency level relating to a 2-week period then an output is generated, by the one or more processors of a computing apparatus, which is a function of the referral output. In the above example, the generated output as a function of the referral output would be endoscopy as a service referral with a 2-week period as the urgency level. The one or more processors of, for example, computing apparatus of Figure 3 then output this generated output, where the output may be sent to the relevant department equipped to deal with the referral. The relevant department may be determined by accessing a library of local services as described below.
[0081] In embodiments, generating an output as a function of the referral output may comprise accessing a library of local services and determining a relevant local. For example, once it is determined that the referral output is endoscopy as a service referral with a 2-week period as the urgency level, then a library of services provided by the hospital or clinic which received the referral (hereafter referred to as the receiving hospital or clinic) may be accessed by a computing apparatus. The library of local services may detail the departments, clinicians and / or services provided by the receiving hospital or clinic. The library of local services may detail the availability of appointments for the departments, clinicians and / or services provided by the local hospital or clinic. The library of local services may be a database stored on a server or the cloud comprising information relating to the departments, clinicians and / or services provided by the receiving hospital or clinic and the availability of appointments for the departments, clinicians and / or services provided by the local hospital or clinic. A computing apparatus such as the one shown in Figure 3 may access the library of local services and then determine the relevant local service.
[0082] For example, if the referral output relates to endoscopy as a service referral which is to be performed within a 2-week period as the urgency level, then the computing apparatus such as the one shown in Figure 3 may access a library of local services relating to the receiving hospital or clinic and determine if the receiving hospital or clinic has the facilities or resources to perform the endoscopy referral within the 2-week period. Once it has been determined that the facilities or resources are available at the receiving hospital or clinic and the relevant local service has been selected, then the computing apparatus generates an output as a function of the referral output, wherein the generated output comprises the referral output and the determined local service. The computing apparatus may then output this generated output, where the output may be transmitted to the selected service within the receiving hospital or clinic by computing apparatus 300.
[0083] Rejecting the referral comprises sending the referral back to the referrer with appropriate feedback, manually reviewing and triaging the referral, where the manual review and triage may be conducted by a clinician or a consultant or redirecting the referral to the department or service as determined by the generated referral output. As an example, a referral is redirected if the GP referral indicates that the patient requires the services provided by the cardiology department i.e., it is a cardiology referral, however, the referral output as determined using clinical guidelines indicates that the patient should in fact be referred to a respiratory specialist then the referral is directed to the department and / or specialist equipped to deal with respiratory referrals.
[0084] The computer-implemented method 100 for referring a patient, as described in Figure 1, is suitable for performance by a computing apparatus such as computing apparatus 300 as shown in Figure 3 and described below.
[0085] Computing apparatus 300 may comprise a computing device, a server, a mobile or portable computer and so on. Computing apparatus 300 may be distributed across multiple connected devices. Other architectures to that shown in Figure 3 may be used as will be appreciated by the skilled person.
[0086] Referring to Figure 3, computing apparatus 300 includes one or more processors 310, one or more memories 320, a number of optional user interfaces such as visual display 330 and virtual or physical keyboard 340, a communications module 350, and optionally a port 360 and optionally a power source 370. Each of components 310, 320, 330, 340, 350, 360, and 370 are interconnected using various busses. Processor 310 can process instructions for execution within the computing apparatus 300, including instructions stored in memory 320, received via communications module 350, or via port 360.
[0087] Memory 320 is for storing data within computing apparatus 300. The one or more memories 320 may include a volatile memory unit or units. The one or more memories may include a non-volatile memory unit or units. In embodiments, the one or more memories 320 may also be another form of computer-readable medium, such as a magnetic or optical disk. One or more memories 320 may provide mass storage for the computing apparatus 300. Instructions for performing a method as described herein may be stored within the one or more memories 320.
[0088] The communications module 350 is suitable for sending and receiving communications between processor 310 and remote systems.
[0089] The port 360 is suitable for receiving, for example, a non-transitory computer readable medium containing one or more instructions to be processed by the processor 310.
[0090] The processor 310 is configured to receive data, access the memory 320, and to act upon instructions received either from said memory 320 or a computer-readable storage medium connected to port 360, from communications module 350 or from user input device 340.
[0091] The computing apparatus 300 may receive, via the communications module 350, a referral from, for example, the e-RS system. The referral may be a GP referral. The referral may comprise referral text and a referrer input. The referral may comprise information relating to the patient such as the patient’s name, age, gender, symptoms and / or medical history (referral text) and information relating to the recommended service and the urgency level (referrer input) as determined by, for example, a GP based on the GP’s diagnosis of the patient.
[0092] The processor 310 may be configured to follow further instructions stored in the memory 320 to extract clinical information from the referral text.
[0093] The processor 310 may be configured to follow further instructions stored in the memory 320 to analyse at least the extracted clinical information from the referral text. The processor 310 may be configured to follow further instructions stored in the memory 320 to compare the extracted clinical information to a database in order to generate a referral output. As an example, the extracted clinical information may be compared to guidelines from a regulatory body such as NICE.
[0094] The processor 310 may be configured to follow further instructions stored in the memory 320 to generate a referral output as a function of at least the analysed extracted clinical information.
[0095] The processor 310 may be configured to follow further instructions stored in the memory 320 to compare the referral output to the referrer input and determine whether or not the referral output corresponds to the referrer input.
[0096] The processor 310 may be configured to follow further instructions stored in the memory 320 wherein if the referral output corresponds to the referrer input, the referrer input is determined to be an approved referrer input, wherein if the referrer input is approved, the processor 310 may be configured to follow further instructions stored in the memory 320 to generate an output as a function of the correspondence of the referral output to the referrer input, wherein the generated output comprises the approved referrer input. The processor 310 may be configured to follow further instructions stored in memory 320 to output the generated output. The processor 310 may be configured to follow further instructions stored in the memory 320 to display the generated output comprising the approved referrer input on the display 330 and / or transmit the approved referrer input to the relevant department equipped to deal with the referral.
[0097] The processor 310 may be configured to follow further instructions stored in the memory 320 to access a library a of local services and determine a relevant local service when the referrer input is approved. The processor 310 may be configured to follow further instructions stored in the memory 320 to generate an output as a function of the correspondence of the referral output to the referrer input, wherein the generated output comprises the approved referrer input and the selected local service. The processor 310 may be configured to follow further instructions stored in the memory 320 to output the generated output comprising the approved referrer input and the selected local service. The processor 310 may be configured to follow further instructions stored in the memory7 320 to display the generated output comprising the approved referrer input and the selected local service on the display 330 and / or transmit the approved referrer input to the selected service.
[0098] The processor 310 may be configured to follow further instructions stored in the memory 320 wherein if the referral output does not correspond to the referrer input, the referrer input is not approved, wherein if the referrer input is not approved, the processor 310 may be configured to follow further instructions stored in the memory 320 to send the referral back to the referrer with appropriate feedback, or send the referral to the department or service as determined by the generated referral output. As an example, if the referrer input as determined by a GP indicates that the patient requires the services provided by the cardiology department i.e., it is a cardiology referral, however, the referral output as determined by a computing apparatus using clinical guidelines indicates that the patient should in fact be referred to a respiratory specialist then the referral is directed to the department and / or specialist equipped to deal with respiratory referrals. The processor 310 may also be configured to follow further instructions stored in the memory 320 to set the referral to the attention of a consultant or clinician for manual review of the referral.
[0099] The computer-implemented method 200 for referring a patient, as described in Figure 2, is also suitable for performance by a computing apparatus such as computing apparatus 300 as shown in Figure 3 and described below.
[0100] The computing apparatus 300 may receive, via the communications module 350, a referral from, for example, the e-RS system. The referral may be a GP referral. The referral may comprise referral text and a referrer input. The referral may comprise information relating to the patient such as the patient’s name, age, gender, symptoms and / or medical history (referral text) and information relating to the recommended service and the urgency level (referrer input) as determined by, for example, a GP based on the GP’s diagnosis of the patient.
[0101] The processor 310 may be configured to follow further instructions stored in the memory 320 to extract clinical information from the referral text.
[0102] The processor 310 may be configured to follow further instructions stored in the memory 320 to analyse at least the extracted clinical information from the referral text. The processor 310 may be configured to follow further instructions stored in the memory 320 to compare the extracted clinical information to a database in order to generate a referral output. As an example, the extracted clinical information may be compared to guidelines from a regulatory body such as NICE.
[0103] The processor 310 may be configured to follow further instructions stored in the memory 320 to generate a referral output as a function of at least the analysed extracted clinical information.
[0104] The processor 310 may be configured to follow further instructions stored in the memory 320 to accept or reject the referral based on the generated referral output.
[0105] The processor 310 may be configured to follow further instructions stored in the memory 320 to accept a referral based on the generated referral output, wherein accepting the referral comprises generating an output as a function of the referral output. The processor 310 may be configured to follow further instructions stored in memory 320 to output the generated output. The processor 310 may be configured to follow further instructions stored in the memory 320 to display the generated output comprising the approved referrer input on the display 330 and / or transmit the approved referrer input to the relevant department equipped to deal with the referral.
[0106] The processor 310 may be configured to follow further instructions stored in the memory 320 to access a library a of local services and determine a relevant local service when the referral is accepted. The processor 310 may be configured to follow further instructions stored in the memory 320 to generate an output as a function of the referral output, wherein the generated output comprises the referral output and the selected local service. The processor 310 may be configured to follow further instructions stored in the memory 320 to output the generated output comprising the referral output and the selected local service. The processor 310 may be configured to follow further instructions stored in the memory 320 to display the generated output comprising the referral output and the selected local service on the display 330 and / or transmit the approved referrer input to the selected service.
[0107] The processor 310 may be configured to follow further instructions stored in the memory 320 to reject a referral, wherein rejecting the referral comprises sending the referral back to the referrer with appropriate feedback, or redirecting the referral to the department or service as determined by the generated referral output. As an example, a referral is redirected if the GP referral indicates that the patient requires the services provided by the cardiology department i.e., it is a cardiology referral, however, the referral output as determined using clinical guidelines indicates that the patient should in fact be referred to a respiratory specialist then the referral is directed to the department and / or specialist equipped to deal with respiratory referrals. The processor 310 may also be configured to follow further instructions stored in the memory 320 to set the referral to the attention of a consultant or clinician for manual review of the referral.
[0108] Based on the above description, computing apparatus 300 can be used for referring a patient. The skilled person would appreciate that other architectures to that shown in Figure 3 may be used.
[0109] It will be appreciated that embodiments of the present invention can be realised in the form of hardware, software or a combination of hardware and software. Any such software may be stored in the form of volatile or non-volatile storage such as, for example, a storage device like a ROM, whether erasable or rewritable or not, or in the form of memory such as, for example, RAM, memory chips, device or integrated circuits or on an optically or magnetically readable medium such as, for example, a CD, DVD, magnetic disk or magnetic tape. It will be appreciated that the storage devices and storage media are embodiments of machine-readable storage that are suitable for storing a program or programs that, when executed, implement embodiments of the present invention. Accordingly, embodiments provide a program comprising code for implementing a system or method as claimed in any preceding claim and a machine-readable storage storing such a program. Still further, embodiments of the present invention may be conveyed electronically via any medium such as a communication signal carried over a wired or wireless connection and embodiments suitably encompass the same.
[0110] Many variations of the method described herein will be apparent to the skilled person. Each feature disclosed in this specification (including any accompanying claims, abstract and drawings), may be replaced by alternative features serving the same, equivalent or similar purpose, unless expressly stated otherwise. Thus, unless expressly stated otherwise, each feature disclosed is one example only of a generic series of equivalent or similar features. The invention is not restricted to the details of any foregoing embodiments. The invention extends to any novel one, or any novel combination, of the features disclosed in this specification (including any accompanying claims, abstract and drawings), or to any novel one, or any novel combination, of the steps of any method or process so disclosed. The claims should not be construed to cover merely the foregoing embodiments, but also any embodiments which fall within the scope of the claims.
[0111] This specification is based on independent research funded by the NHSX ([Artificial Intelligence in Health and Care Award, Referral Intelligence and Triage Automation (RITA), #1810]). The views expressed in this publication are those of the author(s) and not necessarily those of the NHS, NHSX or the Department of Health and Social Care.
Claims
1. A computer-implemented method for referring a patient, the method comprising: receiving a referral from a referrer, the referral comprising referral text and a referrer input;extracting clinical information from the referral text;analysing at least the extracted clinical information from the referral text;generating a referral output as a function of at least the analysed extracted clinical information;comparing the referral output to the referrer input;determining if the referral output corresponds to the referrer input, wherein if the referral output corresponds to the referrer input, the referrer input is approved, wherein if the referrer input is approved:generating an output as a function of the correspondence of the referral output to the referrer input, wherein the generated output comprises the approved referrer input, andoutputting the generated output.
2. A computer-implemented method according to claim 1, wherein generating an output as a function of the correspondence of the referral output to the referrer input comprises accessing a library of local services and determining a relevant local service as a function of the approved referrer input.
3. A computer-implemented method according to any preceding claim, wherein extracting clinical information from the referral text comprises using a language processing technique to extract the clinical information from the referral text.
4. A computer-implemented method according to claim 3, wherein the language processing technique is natural language processing (NLP).
5. A computer-implemented method according to any preceding claim, wherein analysing at least the extracted clinical information from the referral text comprises comparing the extracted clinical information with a database.
6. A computer-implemented method according to claim 5, wherein the database comprises clinical guidelines.
7. A computer-implemented method according to claim 6, wherein the clinical guidelines are from a regulatory body or an independent organisation.
8. A computer-implemented method according to claim 7, wherein the regulatory body is National Institute for Health and Care Excellence (NICE).
9. A computer-implemented method according to any preceding claim, wherein the referrer is a health practitioner, and the referral is made by the health practitioner.
10. A computer-implemented method according to any preceding claim, wherein the referral text comprises patient information, the patient information comprising at least one of: patient’s name, age, gender, symptoms and medical history.
11. A computer-implemented method according to any preceding claim, wherein the referrer input comprises at least one of service referral and / or an urgency level as determined by the GP.
12. A computer-implemented method according to any preceding claim, wherein the referral output comprises at least one of a service referral and / or an urgency level.
13. A computer readable medium having instructions stored thereon, which when executed by a processor, cause the processor to implement a method according to any of claims 1 to 12.
14. A computing apparatus for referring a patient, the apparatus comprising:one or more memory units; andone or more processors configured to execute the instructions stored in the one or more memory units to perform the method of any of claims 1 to 12.