Article Text

Protocol
Variation in daytime general practice services and association with out-of-hours use for older patients: protocol for a Danish cohort study
  1. Jonas K Olsen1,
  2. Sonja Wehberg1,
  3. Frans Boch Waldorff2,
  4. Daniel Pilsgaard Henriksen3,
  5. Jesper Lykkegaard1
  1. 1Research Unit for General Practice, Department of Public Health, University of Southern Denmark, Odense, Denmark
  2. 2Section of General Practice and Research Unit for General Practice, Department of Public Health, University of Copenhagen, Copenhagen, Denmark
  3. 3Department of Clinical Pharmacology, Odense University Hospital, Odense, Denmark
  1. Correspondence to Dr Jonas K Olsen; jkolsen{at}health.sdu.dk

Abstract

Introduction Out-of-hours primary care services cannot provide the same continuity and coordination of care as general practice. Thus, patients with high risk of complex care trajectories should, when possible, be treated by the general practitioner during daytime opening hours. This study aims to analyse the variation among general practices in the frequencies of daytime services for persons aged ≥75 years and how it relates to the patients’ use of out-of-hours services.

Methods and analysis Register-based cohort study of all Danish citizens aged ≥75 years, of whom >98% are listed with a general practice. Using Poisson regression, we will estimate each practice’s excess variation in delivered daytime services compared with the expected based on the characteristics of its listed patients. Delivered daytime services will be analysed overall and separately for face-to-face, phone, email, home visit and preventive services. The association with the use of out-of-hours services will be analysed by Poisson regression.

Ethics and dissemination Complying with European data protection rules, the legal services at University of Southern Denmark (Research & Innovation Organisation) approved the data processing activities regarding this project (journal number 11.593). According to section 14.2 of the Act on Research Ethics Review of Health Research Projects, because the study is based solely on register data, approval from the ethics committee and informed consent are not required. Results from the study will be disseminated as publications in peer-reviewed scientific journals and at international conferences.

  • Aged
  • Primary Care
  • Health Services Accessibility
  • ACCIDENT & EMERGENCY MEDICINE
  • Health Services for the Aged
http://creativecommons.org/licenses/by-nc/4.0/

This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.

Statistics from Altmetric.com

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

STRENGTHS AND LIMITATIONS OF THIS STUDY

  • The study uses high-quality virtually complete nationwide register data.

  • Publishing and adhering to this protocol will prevent the authors from selecting in the coming analyses and results.

  • The observational study design limits our ability to draw conclusions regarding the potential causal relations between provision of daytime services and out-of-hours service use.

Background

Out-of-hours (OOH) primary care services are designed to manage acute conditions that cannot wait until the following workday.1 Low-urgency contacts in the OOH primary care services are a costly use of services compared with daytime general practice,2 and OOH services cannot provide the same, often necessary, continuity and coordination of care as the general practitioner (GP).1 Therefore, OOH services should be used only in case of urgency and not because of lack of convenient access to the GP.

In Denmark, almost all Danish citizens are listed with a specific general practice, who acts as primary care provider and gatekeeper to more extended healthcare services (eg, at hospitals).3 The general practices must serve their listed patients from 08:00 to 16:00 on all workdays.2 OOH services are organised in some regions by a GP cooperative (GPC) available only OOH and in others by a public hospital medical helpline 1813 (MH-1813) available for primary care emergencies round the clock. First contact is always by telephone. Phone triage is conducted by GPs in the GPCs and primarily by nurses in the MH-1813.

Use of OOH services instead of the regular GP is especially problematic for older patients because of the higher frequency of multimorbidity and complex care trajectories.4 Such cases often require higher continuity of care than possible in the OOH services and more often lead to hospitalisation when treated by an unfamiliar doctor.5–7

Older patients usually have a high frequency of contacts to general practice,8 and the use of OOH services increases with ageing.9 However, large variation has been shown in the level of daytime service between general practices.10 For patients with chronic disease, the frequencies of daytime and OOH services are positively associated.11 This is also likely to apply to older patients in general. It is, however, unknown how a change in the delivery of daytime services will affect the use of OOH services.

A sufficient level of daytime services in general practice is likely to reduce the use of OOH services by three mechanisms: (1) by providing timely primary care during the course of acute illness,12 (2) by long-term prevention and treatment of risk factors and progressive diseases,13 and (3) by building a relation with the patient and becoming the usual and confident first access point in the healthcare system.14

The best combination of daytime services for older patients in terms of preventing unnecessary use of OOH services is the situation where fewer than the provided daytime services are associated with increased use of OOH services and more are not associated with a reduction. The level is likely to depend on the types and combinations of services and on patient characteristics such as multimorbidity level,11 socioeconomic status15 and travel distance to OOH primary care services.15

This study aims to analyse the variation among general practices in the frequencies of daytime services for persons aged ≥75 years and how it relates to the patients’ use of OOH services.

Methods

Design

This is a nationwide register-based cohort study.

Ethics

Complying with European data protection rules, the legal services at University of Southern Denmark (Research & Innovation Organisation) approved the data processing activities regarding this project (journal number 11.593). According to section 14.2 of the Act on Research Ethics Review of Health Research Projects and section 10 of the Data Protection Act, because the study is based solely on register data, approval from the ethics committee and informed consent are not required.16 17

Setting

The northwest European country of Denmark has 5.8 million citizens. Most healthcare services including general practice and OOH services are fully tax paid. More than 98% of the population is listed with a specific self-chosen general practice. General practices are privately owned and paid capitation fees (one-third) and fees for services (two-thirds).3 GPs in the GPCs are paid fees for service, while personnel in the MH-1813 are on salary.

Service types in Danish general practice

The GPs electronically report each performed service to the regional health insurance for remuneration. The available services are regular consultation, consultation by phone or email, home visit, phone or email consultation with municipality services, assessment of patients at nursing homes, video conference with the municipality or secondary services, and talk therapy (table 1). In addition to these services, two main preventive services in general practice exist: annual chronic care consultation and preventive home visits. The annual chronic care consultation can be performed once per patient per chronic condition per year. Preventive home visits can be performed once per year to frail older individuals (often ≥75 years with multiple chronic diseases).18 Both services are remunerated about twice the amount for regular consultations and home visits, respectively. In addition, several regional special services exist (§2 services) (table 1). These will be categorised when data are available.

Table 1

Service types and categories in Danish general practice

Population

For each of the years 2017–2022, the study population will include all persons aged ≥75 years who on their birthday were listed with a Danish general practice. The practices must have registered services each month of the current year. Practices with less than 500 listed or 20 eligible patients at the beginning and end of the study will be excluded.

Data sources

All data will be linked at the individual level through each citizen’s unique personal identification number provided by the Danish Civil Registration System.19 Data on types of services in general practice from the Danish National Health Service Register20; data on the link between the patient and the general practice from the Patient List Database; data on sex, age, emigration and death will be obtained from the Danish Civil Registration System19; data on marital status, education, income, ethnicity and home healthcare services from Statistics Denmark21; data on hospital admissions, dates, the International Classification of Diseases 10th revision (ICD-10) diagnosis codes and procedure codes from the Danish National Patient Register22; data on psychiatric hospital admissions and ICD-10 diagnosis codes from the Danish Psychiatric Central Research Register23; and data on redemption dates, redeemed volumes and prescriber of prescribed medications classified according to the Anatomical Therapeutic Chemical (ATC) Classification codes from the Danish National Prescription Register.24

Definition of variables

The Nordic Multimorbidity Index will be used for classifying the patients’ levels of multimorbidity divided into quintiles.25 Ethnicity will be divided into Danish, other western background (European Union, USA, Canada, New Zealand or Australia) and other. Home healthcare services will be divided into five categories: none, lowest, middle and highest tertile, and nursing home resident. The tertiles will be based on allocated time per week by the municipality in minutes. Polypharmacy will be calculated as purchasing more than five different prescription drugs (on fourth level of ATC code) during the past 120 days and categorised into four groups (0, 1–4, 5–9 and 10+) of unique prescription drugs. District will be dichotomised into rural or urban. Travel distance to general practice and hospital will be categorised when data are available. Cohabitation will be divided into single or cohabitating. Household income and wealth will be divided by number of cohabitants in the household and split into quintiles for the study population and used as ordinal categorical values. The composition of the general practices will include the number of GPs divided into 1 (solo practice), 2, 3–4 and 5+, the mean age of the GPs, and the proportion of female GPs and the seniority of GPs both divided into quintiles. For details, see online supplemental material 1.

Statistical analysis

The study will be published in two separate papers. The first focuses on the analysis of variation in services for older patients across general practices. The second explores the association between variation in GP daytime services for all older patients and OOH service use by the listed older patients.

Paper 1: analysis of variation in services for older patients across general practices

This study will provide information on how general practices differ in the level and combination of daytime services for their older patients and to what extent this variation is related to differences in the patients’ health, geographical and socioeconomic factors, and the practices’ organisational characteristics.

The practices’ crude and adjusted rates of services will be analysed using Poisson regression. Four models of adjustment will be developed each including the former model: (1) patient health factors: age, sex, multimorbidity, ethnicity, polypharmacy and level of home healthcare services; (2) patient geographical factors: rural or urban district, travel distance from home address to the general practice, and distance to the appointed emergency department; (3) patient socioeconomic factors: cohabitation, household income and household wealth; and (4) practice factors: GP composition (number, age, gender and seniority). All adjustments will be made on patient level. Separate analyses will be made for each service type, that is, face-to-face services, phone, email, home visits and preventive services (table 1).

Funnel plots will be constructed to display variation between practices in the crude and adjusted rates for overall services and the five service categories.

A subgroup analysis will be made exploring characteristics of general practices outside control limits set at 95% and 99.8% compared with practices within limits.

Using latent profile analysis,26 the practices will be grouped according to the adjusted rates of face-to-face services, phone, email, home visit and preventive services (table 1) defining each practice’s service characteristic.

Paper 2: analysis of the association between variation in practice service characteristics and OOH service use for older patients

This study will investigate how changes in the level and combination of daytime services in general practice are associated with older patients’ use of OOH services.

All patients ≥75 years will be followed 1 year from their birthday in the study year. Their exposure will be the practice’s service characteristic from the year up to their birthday calculated separately for each patient. By use of Poisson regression, we will analyse the association between the practices’ service characteristics and the listed patients’ use of OOH services. The models will consider the overall rate of contacts to OOH services (primary outcome) and rates, respectively, for phone calls, consultations and home visits. Contacts within 1 day of each other will be merged. Four models of adjustment will be developed: (1) crude analysis, (2) adjusted for patient factors, (3) adjusted for practice factors, and (4) adjusted for patient and practice factors. All adjustments will be made on patient level. The practice service characteristic with the lowest use of overall OOH services will be used as reference.

Funnel plots will be constructed to display variation in use of OOH services between practices for each of the models.

Stratified analyses will be performed on low, moderate, and high multimorbidity and socioeconomic status, and on sex, age groups, urbanisation, and practice characteristics.

Sensitivity and power analyses

To assess a further reduction of the risk of confounding by indication, a sensitivity analysis will be made excluding the index patient from the calculation of her/his exposure status from the count of patients and services used to calculate the patient’s practice’s service characteristic.

Analyses of time listed with the general practice before follow-up will be performed comparing patients listed, respectively, 0–1, 2–5 and 5+ years with their current practice.

No statistical power analysis was performed due to having virtually complete data from the Danish registers.

To assess for unmeasured confounding, the E-value will be calculated. The E-value estimates the minimum strength of association an unmeasured confounder would need to have to nullify the estimated association between the exposure and outcome given the covariates.27

Patient and public involvement

A panel of older patients and next of kin will assist interpretation and dissemination of the study results.

Discussion

Key results

The first study will provide information on how general practices differ in the level and combination of daytime service for their older patients and to what extent this variation is related to differences in the included patient population characteristics. The second study will provide insights into if changes in the level or combination of daytime services are associated with older patients’ use of OOH services. The studies will facilitate future research on how variation among general practices’ service characteristics correlates with the prognosis and overall healthcare usage of older patients.

Such knowledge will inform whether and which types of changes in the level of daytime services in general practice are likely to improve prognosis and prevent unbeneficial and maybe even harmful use of healthcare resources. Such changes could be influenced by adjusting fees for services to incentivise beneficial levels or combinations of daytime services.

Strengths and limitations of these studies

The use of high-quality virtually complete nationwide registers reduces the risk of selection bias and increases generalisability. Rich individual data on patients enable adjustment for a plethora of health, geographical and socioeconomic factors isolating the variation excess of differences in patient population in paper one and reducing the risk of confounding in paper two.

By defining patient exposure as the practice delivered daytime services for all patients aged ≥75 years, our method limits the risks of confounding by indication, which is further limited in the sensitivity analysis excluding the index patient from the calculation of his/her exposure status.

Our study has some limitations. Regarding our exposure, two limitations are of notice: first, in general, only one type of contact to general practice can be registered for compensation by the national health insurance per day,28 which may result in a small underestimation of the true level of daytime services for patients contacting their GP multiple times during a single day. Second, during daytime, only telephone services where the patient has been offered advice or treatment are remunerated, not booking of appointments or prescription renewals.28 As these rules are applied nationwide, we expect the potential bias to be distributed rather equally and thus to have little to no effect on our estimates.

Older patients seldom change GP even if they are not satisfied with their current practice.29 30 This ensures most patients will be exposed to the same practice service characteristic throughout the study. Those who do change GP during follow-up are to some degree misclassified by their former practice’s service characteristic introducing undifferentiated variability in the models and increasing the risk of a type II error, that is, a failure to reject the null hypothesis when it is, in fact, false. The same applies to patients listed with practices that change service characteristics during follow-up. We chose to keep the two types of patients in the analyses to limit the risk of selection bias. The proportion of patients changing GP during follow-up will be reported.

We expect our results to be of importance and generalisable to healthcare systems with similar GP listing systems and OOH services, for example, the UK,31 the Netherlands32 and the other Scandinavian countries.33 34

Ethics statements

Patient consent for publication

Acknowledgments

We thank statistician Maria Munch Storsveen for support in applying for data from the national registers.

References

Supplementary materials

  • Supplementary Data

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

Footnotes

  • Contributors JL, DPH, FBW, SW and JO designed the study. JO wrote the first draft of this manuscript. SW, JL and JO decided on the statistical methods. JO and JL revised the manuscript, and SW, DPH and FBW contributed to thorough evaluation of the content. All authors approved the final manuscript.

  • Funding The work is supported by Helsefonden (The Health Foundation) (grant no: 21-B-0148), University of Southern Denmark (grant no: not applicable), Fonden for Almen Praksis (Foundation for General Practice) (grant no: not applicable) and the Danish Regions (grant no: not applicable).

  • Competing interests None declared.

  • Patient and public involvement Patients and/or the public were involved in the design, or conduct, or reporting, or dissemination plans of this research. Refer to the Methods section for further details.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.