Abstract
Background There is a trend towards consolidating smaller primary care practices into larger practices worldwide. However, the effects of practice size on quality of care remain unclear.
Aim This review aims to systematically appraise the effects of practice size on the quality of care in primary care.
Design and setting A systematic review and narrative synthesis of studies examining the relationship between practice size and quality of care in primary care.
Method Quantitative studies that focused on primary care practices or practitioners were identified through PubMed, CINAHL, Embase, Cochrane Library, CRD databases, ProQuest dissertations and theses, conference proceedings, and MedNar databases, as well as the reference lists of included studies. Independent variables were team or list size; outcome variables were measures of clinical processes, clinical outcomes, or patient-reported outcomes. A narrative synthesis of the results was conducted.
Results The database search yielded 371 articles, of which 34 underwent quality assessment, and 17 articles (13 cross-sectional studies) were included. Ten studies examined the association of practice size and clinical processes, but only five found associations of larger practices with selected process measures such as higher specialist referral rates, better adherence to guidelines, higher mammography rates, and better monitoring of haemoglobin A1c. There were mixed results for cytology and pneumococcal coverage. Only one of two studies on clinical outcomes found an effect of larger practices on lower random haemoglobin A1 value. Of the three studies on patient-reported outcomes, smaller practices were consistently found to be associated with satisfaction with access, but evidence was inconsistent for other patient-reported outcomes evaluated.
Conclusion There is limited evidence to support an association between practice size and quality of care in primary care.
INTRODUCTION
Primary care is the first point of contact a person has with the health system,1 and primary care practices can play a vital role in helping patients achieve better quality, better outcomes, better cost effectiveness, and better health status.2 The importance of primary care in delivering high-quality services and care is therefore widely recognised.
Quality of care has been defined as ‘the degree to which health services for individuals and populations increase the likelihood of desired health outcomes and are consistent with current professional knowledge’.3 According to Donabedian, measures of quality of care can be classified into three categories:
structure, which refers to the facilities, personnel, and policies and other organisational characteristics to deliver medical care;
processes, which refer to services provided to or for patients; and
outcomes, which refer to changes in a patient’s health status and patient satisfaction
The organisation that a physician works in is a factor that can influence the quality of medical care. Practice size is an organisational characteristic that may influence the processes and outcomes of care in primary care practices. In recent years, there has been some concern that single-handed or smaller practices do not provide as high a quality of care as group or larger practices.5,6 Further, there has been an increasing trend towards larger primary care practices, owing to the retirement of older GPs from single-handed practices, and interdisciplinary teams to deliver better and more affordable services to patients.2 The number and proportion of group practices have also grown rapidly in the UK and Taiwan.7–9
Some studies have found that larger practices tend to adopt evidence-based care processes, and have the economies of scale to employ staff and effectively use information technologies to implement and assist with quality-improvement processes.5,10 However, smaller practices performed better in facilitating closer relationships of physicians and staff with patients, and were often regarded by patients as being more accessible, achieving higher levels of patient satisfaction.10,11
Apart from the issue of quality, the definition of practice size also varies and is measured differently across countries. The size of primary care practices can range from solo/single-physician practices to group practices. In the US, medical groups are defined as organisations with three or more physicians; independent practice associations (IPAs) are organisations that contract with health maintenance organisations representing large numbers of medical groups and one- and two-physician practices; and hybrids refer to organisations that consist of a medical group and an IPA.10 In England, practice size is either defined by the number of patients who choose to register with a general practice (practice list), or by the number of GPs, ranging from single-handed GPs to multi-partner practices.8
How this fits in
There is concern that smaller practices do not provide as high a quality of care as larger practices. A systematic review and narrative synthesis of 13 studies shows that there is limited evidence to support an association between practice size and quality of care in primary care. Practice size is not the only organisational characteristic that can influence the quality of medical care. Future research should be done to ascertain if other organisational factors such as team climate, financial incentives, and time spent with patient, may have greater impact on the quality of care in the primary care setting.
Given the quandary surrounding the definition and quality of care provided by smaller practices, the evidence on the effects of practice size on clinical outcomes has not been conclusive. The aim of this study was to systematically review the effect of practice size on the quality of care in primary care settings.
METHOD
Figure 1 presents a flowchart of the article search and selection process. A search for quantitative studies that examined the relationship between practice size and quality of care in primary care practices or among primary care practitioners (GPs, general internists, paediatricians, family physicians, and obstetrician/gynaecologists) was performed in PubMed, CINAHL, Embase, Cochrane Central Register of Controlled Trials (CENTRAL), CRD databases, ProQuest dissertations and theses, conference proceedings, and MedNar databases. Studies in English published from January 1990 to December 2010 were considered.
Figure 1 Flowchart of article search and selection process.
A search string of MeSH and non-MeSH terms related to practice size and quality of care were identified, and undertaken for all included databases. Practice size could be measured in terms of practice list size, team size, or practice partnerships, that is, whether or not a practice is in partnership with other practices. Quality of care was defined as any objective measure of clinical process, clinical outcome, or patient-reported outcome, which was data reported directly by the patient, and includes functional status, health-related quality of life, satisfaction with treatment, and treatment adherence.12
The titles and abstracts of studies identified from the database search were screened, and full-text copies of studies that met the inclusion criteria (that is, participants, intervention, and outcome measures) were retrieved. If it was unclear from the abstract whether the study met the inclusion criteria, the full-text copy was also retrieved. Two reviewers independently assessed the articles, and disagreements were resolved through discussion. On selection, the reference lists of the studies were searched for additional studies. All selected studies then underwent a quality assessment using a modified critical appraisal checklist from the Joanna Briggs Institute (JBI) critical appraisal checklist for descriptive/case series studies.13
Data were extracted using a tool modified from that developed by the JBI.13 A narrative synthesis of the results was conducted, as a pooled statistical analysis was not possible, owing to the heterogeneity between the studies. In most of the studies, odds ratios were calculated comparing large practices to small practices as the reference group. However, in a few studies, the odds ratios were calculated using large practices as the reference group. For consistent interpretation across all studies, the odds ratios reported in the latter studies were inverted.
RESULTS
Detailed tables of the results are available in a full systematic review from the JBI Library of Systematic Reviews.14
Description of studies
A total of 371 articles were identified from the database search. After reviewing the titles and abstracts, 303 articles were excluded and the remaining 68 articles were retrieved for a detailed examination. Thirty-eight articles were further excluded, yielding 30 relevant articles. The reference lists of the 30 relevant articles were searched, and an additional four articles were identified. All 34 articles underwent a methodological quality assessment, after which 17 articles were excluded, leaving 17 articles to be included in the review (Figure 1).
The 17 articles originated from 13 cross-sectional studies. Two articles were separate publications of one study,15,16 three articles were separate publications of a second study,17–19 and another two articles were separate publications of a third study.20,21 Nine studies were conducted in the UK,15–27 two in the US,28,29 one in Canada,30 and one in the Netherlands.31 Six of the 13 studies focused on primary care practices,15–19,22,24,25,27 while four looked at primary care practitioners,28–31 and three examined both primary care practices and practitioners.20,21,23,26
In three studies, practice size was measured as a continuous variable of list size in thousands,20,21,24,25 and in another three studies, practice size was measured as a dichotomous variable comparing single-handed practices against group practices.22,27,29 Practice size was measured as a categorical variable comparing practices grouped into different sizes in three of the studies,17–19,26,28 while four other studies used more than one measure of practice size.15,16,23,30,31 Of the 13 studies, eight examined the effects of practice size on clinical processes,17–19,22,23,27–31 one on clinical outcomes,26 and two on patient-reported outcomes.20,21,24 One study looked at both clinical processes and outcomes,25 and another study looked at clinical processes and patient-reported outcomes.15,16
Effect findings
The included studies used different levels of significance in the reporting of results. In this review, independent variables with P<0.05 were reported as significant.
Clinical processes
Table 1 presents the results of the studies that examined the effects of practice size on clinical processes. The measures of clinical processes varied between studies and included the use of diagnostic tests, recording of symptoms, recording of risk factors, medication prescription, patient education, specialist referrals, timely review, adherence to clinical practice guidelines, preventive services, and the management of diseases. Ten of the 13 studies reported effects of practice size on clinical processes.15–19,22,23,25,27–31
Table 1 Measures of clinical processes
Of the 10 process measures covered across the studies, only four had significant associations with practice size. Larger practices were found to have higher specialist referral rates for an eating disorder service,23 were more likely to adhere to clinical practice guidelines for children with attention deficit/hyperactivity disorder (ADHD),29 to provide better diabetes management,15,16 and to deliver preventive services such as cytology targets, pneumococcal vaccination, mammography, and haemoglobin A1c testing.22,27,28
Clinical outcomes
Two of the 13 studies examined the effects of practice size on clinical outcomes.25,26Table 2 presents the results of the studies. One study found that patients with diabetes in larger practices tended to have lower random haemoglobin A1 value,26 while the other study found no significant association between practice size and clinical outcomes in patients with ischaemic heart disease (IHD).25
Table 2 Measures of clinical outcomes
Patient-reported outcomes
Three studies examined the effects of practice size on patient-reported outcomes (Table 3).15,16,20,21,24 All three studies demonstrated that larger practices had significantly lower scores for access and treatment by receptionists,15,16 decreased surgery and consultation satisfaction,20,21 and decreased satisfaction with getting through on the phone, being able to get an appointment on the same day or in the next 2 days, being able to get an appointment >2 days in advance, being able to get an appointment with a particular GP, and the hours the GP surgery was open.24
Table 3 Patient reported-outcomes
DISCUSSION
Summary
This review examined the effects of practice size on the quality of care in the primary care setting. Owing to the heterogeneity across the studies in definitions of practice size, types of independent variables, and outcome measures, the findings have to be interpreted with caution in view of the methodological limitations and external validity of the studies.
The higher specialist referral rates for eating disorders in larger practices may be a result of the greater number of sympathetic professionals available.23 Also, larger practices may be more likely to incorporate guidelines for the diagnosis and treatment of ADHD, as solo practices may have poorer access to mental health services, limited insurance coverage, and other potential system barriers to the delivery of ADHD care.29 The association of larger practices with better diabetes management also suggests that there may be positive effects of having a multidisciplinary team,15,16 that shared care with nurses allowed for certain processes to be effectively performed by nurses,32 or that larger practices simply have better information technologies to facilitate execution of clinical processes.10
The mixed results regarding the effects of larger practice sizes on preventive services could be due to methodological limitations, different settings, and varying independent variables and process measures across the studies. Although larger practices appeared to have performed better in certain clinical processes such as higher specialist referral rates for eating disorder and greater adherence to guidelines for the diagnosis and treatment of ADHD, these clinical processes were specific to the study and disease context, and, given that evidence was limited and inconsistent, it would be difficult to make any meaningful generalisations regarding the association between practice size and clinical processes.
Of the two studies that investigated the effect of practice size on clinical outcomes, one study found a significant association between larger practices and lower random haemoglobin A1 values in patients with diabetes while the other found no association between practice size and the achievement of optimal blood pressure, cholesterol, and body mass index readings in patients with IHD. The contrasting result could be due to outcome measures being less sensitive to changes in quality of care as compared to process measures.33
For patient-reported outcomes, smaller practices were consistently found to be associated with better access to care and better patient satisfaction.15,16,20,21,24 This may be because patients were more familiar with the few receptionists in smaller practices as compared to the many receptionists in larger practices.21 Moreover, smaller practices are likely to have a lower volume of calls compared to larger practices, and therefore maintain a better ratio of administrative staff to calls, which would improve satisfaction with telephone access.24 Smaller practices also tend to have fewer changes of doctors and more personal service,20 compared to larger practices where the length of time spent during consultation is limited because of the large volume of patients. Nonetheless, although there was some evidence for smaller practices having better patient-reported outcomes, this was limited to certain outcomes such as access, and is specific to the population and context.
Strengths and limitations
There are limitations to the interpretation of this review. First, not all relevant studies may have been identified, as only studies in the English language from 1990 to 2010 were included in the search strategy. Secondly, the heterogeneity of the studies did not allow for a quantitative meta-analysis. Nonetheless, the estimated effects and confidence intervals have been reported to provide an assessment of the magnitude and meaning of the relationships between practice size and the various outcome variables. Lastly, there is a huge variation across countries in the way primary care is being delivered to patients, and also in the definition and measure of practice size. What one country may consider as a small practice may be viewed as a large practice in another country. Similarly, while one country measures practice size according to list size, another country defines practice size by the team size. Hence, it may be inappropriate to make global judgments about care in general practice from these data.
Comparison with existing literature
Although there was some evidence of larger practices performing better in certain process measures, and smaller practices performing better in selected patient-reported outcomes, the findings lacked consistency across the studies. Nonetheless, this suggests that the relationship between practice size and quality of care is not a simple one. Different types of practices may have different strengths.16 In a review of 30 studies from the UK and US on the relationship between performance and size of primary care groups/trusts (defined as serving about 100 000 people), it was found that increased population sizes beyond 100 000 did not generate cost savings or improvements in overall performance.34 Other studies have suggested that a practice size of 10 000 to 15 000 patients may be optimal for the average physician workload.35 Ultimately, the optimal size of practices is dependent on the aims, functions, and tasks of the organisation, and there may not be one size that would suit all.34
Practice size is also not the only organisational characteristic that can influence the quality of medical care. Other organisational factors such as team climate, financial incentives, and time spent with the patient, may have a greater impact on the quality of care. Future research on these factors should be done to ascertain whether they are better predictors for quality of care in the primary care setting.
Implications for clinical practice
This review provides an updated synthesis of the evidence on the effects of practice size on the quality of care in primary care.36 Only four of the 10 measures of clinical processes and one of two measures of clinical outcomes were significantly associated with larger practices. For patient-reported outcomes, there was some evidence that smaller practices had better patient-reported access, and higher patient satisfaction with surgery and consultations, and with treatment by receptionists. The findings of the review suggest that there is limited evidence available to support an association between practice size and the quality of care in primary care settings.
Acknowledgments
The authors would like to thank Ms Tan Woan Shin and Ms Anusha Govinda Raj for their comments on previous drafts of the review.
Notes
Provenance
Freely submitted; externally peer reviewed.
Competing interests
The authors have declared no competing interests.
- Received January 17, 2013.
- Revision received January 30, 2013.
- Accepted May 7, 2013.
- © British Journal of General Practice 2013
