The Use of a Monitoring Device by General Practitioners During Out-of-hours Care

The Use of a Monitoring Device by General Practitioners During Out-of-hours Care: Protocol of a Prospective Randomised Controlled Trial

All calls that end up on the out-of-hours general practitioners' service (OHGPS), which contain a demand for an urgent home visit, are passed on to the on-call general practitioner (GP). These calls are randomized into two arms: after the patient's informed consent, they are assigned either to one arm where the monitoring device, PICO, is applied together with the GP's general care or to the other arm where only the usual care is provided. All data such as suspected diagnosis, treatment or referral, influence of the parameters, ECG and/or alarms on the management and the user-friendliness are recorded. After 30 days, the diagnosis and evolution is requested from the patient's own GP or, if referred to a hospital, in the hospital in order to be able to compare the effect of the approach by the GP between both arms. The aim is to investigate if 1/ the use of the PICO monitoring device could improve GPs' decisions to refer to hospital or not in urgent cases; 2/ there is a difference between the diagnosis with and without the use of the monitoring device using the final diagnosis by the electronic health record of the own GP of the patient; 3/ the call to send a GP for an emergency contained sufficient information for the OHGPS phone operator to take an appropriate decision; 4/ the build-in alarms help the GP during his intervention; 5/ the PICO is easy to use during an emergency; 6/ the use of the device makes them feel more confident in transmitting the information to the Medical Emergency Team.

INTRODUCTION Calls to an out-of-hours general practitioner's service (OHGPS) for an emergency could originate from different types of callers. Based on the information provided over the phone, it is sometimes difficult to figure out to what extent the call can actually be considered urgent. The development of the out-of-hours services by GPs in Belgium recently led to the introduction of a uniform urgent phone number 1733 for out-of-hours calls. The phone operator can send a GP for an urgent home-visit, ask the patient to go to the OHGPS, to wait until the next available appointment with his/her regular GP or go to the emergency department (ED) immediately. The OHGPS phone operator must be able to differentiate an urgent situation from the vast majority of semi-urgent calls. The assessment of an acutely ill patient relies on the clinical data observed and measured by GPs. During an urgent home visit, a GP is typically multitasking: taking clinical history, performing clinical examination, formulating a diagnostic hypothesis, administering medication and also checking as many vital parameters as possible. There is hardly any time left to make notes. They could also miss some changes in the patient's status and they do not have a monitoring device similar to the one EPs carry. The possibility of checking the monitored records retrospectively, is not only useful to review data during the period between the GP's and the emergency physician's (EP) arrival. Data can also be reproduced at a later stage, e.g. during transport of the patient or during stay at the ED. All information, recorded by a GP before the arrival of the EP, is crucial to support diagnosis, treatment and prognosis. After an extensive literature review, the investigators did not find a single paper on the use of portable monitoring devices by GPs in out-of-hospital emergencies. There are many devices on the market capable of measuring more than one parameter. Unfortunately, most of them are too large and heavy, and do not record all parameters, are unable to transmit results to another device, or do not contain a reproducible memory. A monitoring device, called PICO, was found: it is portable, fits in a doctor's bag and immediately shows all parameters and curves on the display, similar to the EDs' monitors. The data on this device can also be transferred on-site to a computer in which retrospective review of the parameters is possible. All data are saved on both devices. Although this device is not yet commercialised, the future aim is to transfer the data to any device such as a tablet or a cell phone and by all possible means of communication to the prehospital or ED staff. METHOD Study Design Whenever a patient requires medical help, his family or a bystander can call the 112 emergency number or, if a GP is needed, the 1733 number. The urgent calls received by the OHGPS phone operator are transferred to the GP on duty. For the other calls, the patient is invited to go to the OHGPS. For home visits, GPs have a dedicated car with a chauffeur at their disposal, to avoid parking problems and to have additional support available. For each home visit, GPs on duty register all visits, urgent and not urgent, as they usual do. Because they have to write a report to the patient's GP, presenting complaint(s) and diagnosis are recorded for every home visit. Before the study, all GPs and chauffeurs are informed about the study design and the detailed procedures, and have to sign the informed consent form (ICF) respectively for GPs and chauffeurs. The main investigator developed an extensive training video showing how to use the PICO monitor for all GPs and chauffeurs. Each patient seen at home is considered as having requested an urgent visit. Before entering the residence, each patient will be allocated to one of two arms, according to the results of the randomisation process (see statistics section). To be able to include the patient, the GP needs to obtain an ICF from the patient or his legal representative or when the requirements of the World Medical Association's (WMA) Declaration of Helsinki are fulfilled (articles 29 and 30). The reasons for including a patient if no IC can be obtained and if there is no legal representative present, are: a comatose patient, patients with stroke, patients unable to speak, patients in critical situations in which the GP does not have sufficient time to obtain the informed consent. E.g. acute heart failure (HF) with acute pulmonary oedema, patients with hallucinations or confusion from any kind of origin, patients who give oral informed consent but afterwards suffer of cardiac arrest (resulting in the GP to initiate CPR). In all other cases, when taking an informed consent is deemed not feasible, the reason will be recorded on the patient's ICF. Once the patient can be included, the GP performs his clinical tasks. The chauffeur applies the PICO device according to the randomisation process and checks if all parameters and curves are correctly recorded. The GP can follow the evolution of these measurements and both the GP and chauffeur will be alarmed if the device records any abnormal parameter result in a patient. Immediately after the home visit, the GP completes an 'intervention form' as usual, containing the reason of the call, the probable diagnosis, decision and treatment. Additionally, the self-recorded parameters, the use of the monitoring device (yes or no), the level of urgency, the confirmation that an informed consent is obtained, the user-friendliness of the device, its utility and if the use had any influence on their intervention and/or decision, will be recorded, as well as how the communication with the healthcare professionals went. Thirty days after the intervention, the diagnosis of each case will be retrieved from the patients' regular GP's health record and, if the patient was admitted, from hospital. The monitoring device The PICO is a stand-alone patient monitor displaying 0-2 curves and/or 0-6 measured or calculated parameters which can be displayed simultaneously while the monitoring function is working. The device can be operated by using its touchscreen and buttons and has a 3-level alarm system that indicates whenever a parameter is out of the set limits. The basic functions of the monitor (ECG, respiration, temperature and SpO2) are preconfigured. However, external measuring modules can be connected. Respiration is measured by measuring impedance (from the ECG electrodes), SpO2 and pulse rate through a pulse-oximeter and temperature with one channel measurement. The 3-channel ECG displays heart rate (HR) and arrhythmia analysis. Systolic and diastolic parameters of blood pressure can be entered different times manually. With the Bluetooth network connectivity all data can be transferred to a PC or tablet on which a special developed software shows the parameters and curves as on an ED monitoring screen. Statistics Randomization will be stratified at the OHGPS level and by GP's intervention vehicle and will be randomised in blocks of 6. The OHGPS is open on weekends, from Friday 19:00 h until Monday 08:00 h. Based on prevalence of emergency diagnoses in the Intego data, the investigator calculated that 10% of the calls are very urgent and 25% require an urgent decision but are not always followed by a hospitalisation. The presumption is that GPs generally refer 21% of the patients inappropriately to hospital or to patient's own GP. One will need 866 patients, 433 patients in each arm. Data recorded: the number of calls at the 112 dispatch centre by means of the 112 or 1733 dial number and the 112's decision. the number of calls dispatched from 112 to the OHGPS phone operator. the decision of the OHGPS phone operator. all data on the intervention form for both study arms, for each patient, who gave IC. These data consist of those filled in by the GP on-call immediately after the intervention: first (when deciding on referral) and final (when completing the intervention form) diagnostic hypothesis; referral to hospital or not; parameters and curves inspected and results; if any alarm was given and noticed; the impact of the device parameters on the decision; if the device brought new elements; the feeling of the GP about the improvement or not by the device in communication with the healthcare professional; which parameters helped the communication and if they were showed of viewed by the EP. After 30 days the information provided by the patient's own regular GP, regarding the patient's outcome; the information concerning the diagnosis at arrival at the ED and, if available, the outcome of the patient until day 30 of the hospital admission. Analysis The investigator will compare the amount of (verified or experienced) emergencies with the non-urgent home visits, the number of patients correctly referred to hospital or not. The difference in proportions in both groups will be tested by the Z-test. The investigator will also test for the difference in proportions in the groups using the monitoring device and the usual care group, with the z-statistic for two-sample test of proportions reporting the z-test statistic and the associated p-values. All p-values smaller than 0.05 will be considered statistically significant. The statistical analyses will be performed with R software version 3.5.1. The principal investigator will code the recorded diagnoses of the home visits according the International Classification of Primary Care (ICPC). He will compare the diagnoses made during home visits with the overall amount of diagnosis in the Intego registry of diseases in the Flemish population of Belgium to assess generalisability. He will compare for each patient and for the whole study the reason(s) for encounter with the first and final diagnostic hypothesis of the GP and the number of patients referred to an hospital, correctly and incorrectly, and by which mode of transportation they are transported to hospital. All written comments will be coded by two independent researchers and, after agreement, conclusions will be stipulated. The quantitative data will be analysed with MedCalc and R software version 3.5.1 for differences between both arms in proportions of patients correctly referred to hospital or not. For the qualitative data, the answers using the framework method will be coded and conclusions of both investigators compared using kappa statistics. For missing data, the number of missing values for each variable of interest will be reported, the reasons for missing data described and how many individuals are excluded will be indicated. Important differences between individuals with complete and incomplete data will be clarified. If appropriate and if the assumption that missing data was at random has been met, multiple imputation based will be attempted. The intention is to analyse the study with the Intention-To-Treat (ITT) principle, including all patients who gave their informed consent, ignoring what happened to the patient afterwards. Expected duration of trial The study will last for six months, depending on the number of recruited patients with a minimum of 866 cases where an ICF is obtained or when article 29 and 30 of the WMA Declaration of Helsinki is applicable. Quality assurance The device, InnoCare PICO, Type IMH-8M is a portable colour patient monitor and developed, manufactured and distributed by Innomed Medical Inc. and labelled CE 0120. Ethics approval The study was approved by the ethical committee of KU Leuven under the S-number S63046 on 2020, October 07. DISCUSSION The feasibility of the study was confirmed and the responsible persons of the OHGPS agreed on the study. Randomisation is done in blocs of six because two GPs are involved during the same periods. The alarms of the PICO will be set at the same level as the values of the Early Warning Score.

Emergencies, Heart Failure, Pulmonary Embolism, Acute Exacerbation of Chronic Obstructive Pulmonary Disease, Asthma, Acute Coronary Syndrome, Acute Myocardial Infarction, Atrial Fibrillation, Atrial Flutter, Hyperventilation, Pneumonia, Pneumothorax, Cardiac Arrest, Pulmonary Edema, Rhythm; Abnormal

general practioner, monitoring device, vital parameters, ECG, out-of-hours care, RCT, Family practice

The aim is to record if the outcome (referral or not) is influenced by the results of the parameters and/or the ECG.

The number of correct referred participants in the intervention arm is a minimum of 11% higher than the correct referred participants in the usual care arm.

A correct referral is defined as follows: a patient referred to hospital staying more than 12 hours in the hospital; a patient referred only for a diagnostic test, e.g. an X-Ray, an ECG, a blood test, advice of a consultant, and no hospitalisation followed; a patient referred to his own GP was not hospitalised within 72 hours for the same reason. The aim of the study is to investigate the percentage of correct referrals in both arms and to determine if the percentage in the intervention arm is 11% or more higher than in the usual care arm.

The rate of correct diagnosis in the intervention arm is a minimum of 11% higher than the rate of correct diagnosis in the usual care arm.

In the intervention form the GP on duty records the diagnosis. These diagnoses will be compared with the records of the final diagnosis in the electronic health record of the own GP when the patient is sent home or, when the patient is referred to the hospital in the final diagnosis made by the emergency department physician at arrival and after 30 days. The difference between correct diagnoses in the intervention arm should be a minimum of 11 % higher than in the usual care arm.

Do the reasons for encounter in the message from the OGHPS telephone operator to the on-call GP correspond to the severity of the condition (urgent versus non-urgent) as determined by the on-call GP at the patient's location

The reason for encounter which is given to the on-call GP to go for a home visit should point into the direction of an urgent home visit. This message is the one that the 112-1733 dispatcher has passed on to the OHGPS phone operator. The on-call GP then determines on the basis of the patient's complaints whether these complaints indeed correspond to the reason for requesting a home visit. By comparing these reasons, it is possible to establish a better triage protocol both for the OHGPS phone operator and for the 112-1733 dispatcher.

Do the PICO(TM) alarms set for SpO2 assist the GP when the upper or lower limit is exceeded in whether or not to refer the patient to a hospital?

The alarms of the PICO(TM) are set for SpO2 at ≤ 92%, the international cut-off below which one can expect a worse outcome and that a referral to the hospital is strongly recommended if the value is below 90%. The on-call GP records on the intervention form if the alarm helped him/her in the approach of the patient by increasing or decreasing the level of urgency.

Do the PICO(TM) alarms set for the heart rate assist the GP when the upper or lower limit is exceeded in whether or not to refer the patient to a hospital?

The alarms of the PICO(TM) are set at the same upper and lower limits according to the limits of the Early Warning Score (EWS). For the heart rate (normal limits = 51-100 beats per minute (bpm)) the upper and lower limits are resp. <40 bpm and > 130 bpm. Once these limits are exceeded, one can expect a worse outcome and a referral to the hospital is strongly recommended. The on-call GP indicates on the intervention form if the alarm helped him/her in the approach of the patient by increasing or decreasing the level of urgency.

Do the PICO(TM) alarms set for the respiration rate assist the GP when the upper or lower limit is exceeded in whether or not to refer the patient to a hospital?

The alarms of the PICO(TM) are set at the same upper and lower limits according to the limits of the Early Warning Score (EWS). For the respiration rate (normal limits are 9-14 bpm) the upper limit is ≥30 bpm, the lower limit is <9 bpm. Once these limits are exceeded, one can expect a worse outcome and a referral to the hospital is strongly recommended. The on-call GP indicates on the intervention form if the alarm helped him/her in the approach of the patient by increasing or decreasing the level of urgency.

Inclusion Criteria: From the calls at the 112-1733 dispatch sent to the OHGPS dispatch centre, those for whom the decision by the OHGPS dispatcher is to send a GP for an urgent home visit. All GPs on duty and chauffeurs of the OHGPS in Belgium, present during the study period are recruited after signing an ICF. GPs' age, gender and years of practice will be recorded. All patients 18 years and older, for whom a home visit is requested, seen by a participating GP and if the informed consent form is signed either by the patient or by the legal representative, either onsite or at a later time. Exclusion Criteria: Patients younger than 18 years Patients not seen during home visits, Patients failed to provide informed consent Patients with an acute trauma but not in a possible life-threatening situation (e.g. a broken bone) Victims found lying on the street Patients seen after the intervention of an ambulance, a Primary Intervention Team (PIT) or a Medical Emergency Team (MET) Patients not meeting inclusion criteria Patients refusing to participate

Van den Bruel A, Thompson M, Buntinx F, Mant D. Clinicians' gut feeling about serious infections in children: observational study. BMJ. 2012 Sep 25;345:e6144. doi: 10.1136/bmj.e6144.

Schols AMR, Dinant GJ, Hopstaken R, Price CP, Kusters R, Cals JWL. International definition of a point-of-care test in family practice: a modified e-Delphi procedure. Fam Pract. 2018 Jul 23;35(4):475-480. doi: 10.1093/fampra/cmx134.

Dalbak LG, Straand J, Melbye H. Should pulse oximetry be included in GPs' assessment of patients with obstructive lung disease? Scand J Prim Health Care. 2015;33(4):305-10. doi: 10.3109/02813432.2015.1117283. Epub 2015 Dec 11.

Renier W, Geelen M, Steverlynck L, Wauters J, Aertgeerts B, Verbakel J, Vanbrabant P, Gillet JB, Sabbe M, Buntinx F. Can the heartscan be used for diagnosis and monitoring of emergencies in general practice? Acta Cardiol. 2012 Oct;67(5):525-31. doi: 10.1080/ac.67.5.2174126.

Zicari AM, Marzo G, Rugiano A, Celani C, Carbone MP, Tecco S, Duse M. Habitual snoring and atopic state: correlations with respiratory function and teeth occlusion. BMC Pediatr. 2012 Nov 7;12:175. doi: 10.1186/1471-2431-12-175.

Sieber A, L'Abbate A, Kuch B, Wagner M, Benassi A, Passera M, Bedini R. Advanced instrumentation for research in diving and hyperbaric medicine. Undersea Hyperb Med. 2010 Sep-Oct;37(5):259-69. Erratum In: Undersea Hyperb Med. 2011 Jan-Feb;38(1):83.

Park MH, de Asmundis C, Chierchia GB, Sarkozy A, Benatar A, Brugada P. First experience of monitoring with cardiac event recorder electrocardiography Omron system in childhood population for sporadic, potentially arrhythmia-related symptoms. Europace. 2011 Sep;13(9):1335-9. doi: 10.1093/europace/eur159. Epub 2011 May 26.

Hochstadt A, Chorin E, Viskin S, Schwartz AL, Lubman N, Rosso R. Continuous heart rate monitoring for automatic detection of atrial fibrillation with novel bio-sensing technology. J Electrocardiol. 2019 Jan-Feb;52:23-27. doi: 10.1016/j.jelectrocard.2018.10.096. Epub 2018 Nov 1.

Flynn D, Francis R, Robalino S, Lally J, Snooks H, Rodgers H, McClelland G, Ford GA, Price C. A review of enhanced paramedic roles during and after hospital handover of stroke, myocardial infarction and trauma patients. BMC Emerg Med. 2017 Feb 23;17(1):5. doi: 10.1186/s12873-017-0118-5.

Byrne AJ, Jones JG. Responses to simulated anaesthetic emergencies by anaesthetists with different durations of clinical experience. Br J Anaesth. 1997 May;78(5):553-6. doi: 10.1093/bja/78.5.553.

Hansen MB, Lippert FK, Rasmussen LS, Nielsen AM. Systematic downloading and analysis of data from automated external defibrillators used in out-of-hospital cardiac arrest. Resuscitation. 2014 Dec;85(12):1681-5. doi: 10.1016/j.resuscitation.2014.08.038. Epub 2014 Oct 2.

Gale NK, Heath G, Cameron E, Rashid S, Redwood S. Using the framework method for the analysis of qualitative data in multi-disciplinary health research. BMC Med Res Methodol. 2013 Sep 18;13:117. doi: 10.1186/1471-2288-13-117.

Sterne JA, White IR, Carlin JB, Spratt M, Royston P, Kenward MG, Wood AM, Carpenter JR. Multiple imputation for missing data in epidemiological and clinical research: potential and pitfalls. BMJ. 2009 Jun 29;338:b2393. doi: 10.1136/bmj.b2393.

Bagnasco A, Costa A, Catania G, Zanini M, Ghirotto L, Timmins F, Sasso L. Improving the quality of communication during handover in a Paediatric Emergency Department: a qualitative pilot study. J Prev Med Hyg. 2019 Sep 30;60(3):E219-E225. doi: 10.15167/2421-4248/jpmh2019.60.3.1042. eCollection 2019 Sep.

https://www.health.belgium.be/sites/default/files/uploads/fields/fpshealth_theme_file/analyse_2010_2018_nl.pdf

https://www.wma.net/policies-post/wma-declaration-of-helsinki-ethical-principles-for-medical-research-involving-human-subjects/

Core Team (2020). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria.

https://www.huisartsnu.be/2010/nr5/onderzoek/pulsoxymetriescreening-de-huisartsenpraktijk-heeft-het-een-invloed-op-het-beleid

Article in Dutch on pulse oximetry screening in general practice: Renier W; Buntinx F. Pulsoximetriescreening in de huisartsenpraktijk. Heeft het invloed op het beleid? Huisarts Nu 2010;39:190-204.

Article in Dutch on pulse oximetry: indications and observations in case of COPD - Schermer T, Leenders J, In 't Veen H, Van den Bosch W, Wissink A, Smeele I, Chavannes N. Pulsoximetrie: indicaties en observaties bij COPD. Huisarts Wet 2010;53(6):30

Study in British Medical Journal (BMJ) Innovations: 15. Renier WS, Erard K, Sabbe M, Hubloue I, Verbakel JY, Aertgeerts B, Buntinx F. Analytical accuracy of the handheld PICO monitoring device during emergencies. BMJ Innovations 2019;5:20-27.