Clinical Visit Frequencies in Rheumatology – A Systematic Literature Review

Abstract

Introduction

Clinical visits are a fundamental aspect of rheumatic disease care, but recommendations for appropriate visit frequencies are largely absent from guidelines, scarcely studied, and inconsistently reported. The aim of this systematic review was to summarize the evidence pertaining to visit frequencies for major rheumatic diseases.

Methods

This systematic review was conducted according to PRISMA guidelines. Title/abstract screening, full-text screening, and extraction was carried out by two independent authors. Annual visit frequencies were either extracted or calculated and stratified by disease type and country of study. Weighted mean annual visit frequencies were calculated.

Results

273 relevant manuscript records were screened, and 28 were included after applying selection criteria. The included studies were equally divided between US and non-US and were published between 1985 and 2021. Most (n = 16) focused on rheumatoid arthritis (RA), systemic lupus erythematosus (SLE, n = 5), and fibromyalgia (FM, n = 4). For RA, the average annual visit frequencies were 5.25 for US rheumatologists, 4.80 for US non-rheumatologist, 3.29 for non-US rheumatologists, and 2.74 for non-US non-rheumatologists. For SLE, annual visit frequencies for non-rheumatologists were much higher than US rheumatologists (12.3 vs 3.24). For FM, annual visit frequencies were 1.80 for US rheumatologists and 0.40 for non-US rheumatologists. There is a decreasing trend of visit frequency to rheumatologists from 1982 to 2019.

Conclusion

Evidence for rheumatology clinical visits were limited and heterogenous on a global scale. However, general trends suggest more frequent visits in the US and less frequent visits in recent years.

INTRODUCTION

Rheumatology is mainly an outpatient specialty, with patients’ clinical visits often comprising a major aspect of care¹. Timely initiation and sustained maintenance of medical interventions, which typically occur at clinical visits, are associated with improved prognosis and symptomatic amelioration in many rheumatic diseases². However, the appropriate frequency to visit a physician for rheumatic disease follow-up has not been a major focus of study³, resulting in a lack of guidance for rheumatic disease follow-up for common conditions like rheumatoid arthritis (RA)⁴ and osteoarthritis (OA)⁵.

Competing and sometimes contradictory evidence abounds in rheumatology health service literature with regards to optimal visit frequencies. A US-based study from the late 1990s reports more frequent visits to rheumatologists were associated with greater improvements in pain and functional disability RA.⁶ While, a 2021 Danish randomized control trial (RCT) found no difference in patient prognosis between usual care and an experimental arm that featured fewer visits.⁷ An analysis of a US national patient database revealed significant regional variation regarding rheumatology visits and referral patterns in the US; more frequent referral and visit to a rheumatologist was not associated with a patient’s overall satisfaction or perceived health status⁸. After the advent of the COVID-19 pandemic, questions have been raised for both the traditional interval of patient care and the proper role of virtual care moving forward⁹. As the global burden of rheumatic diseases continues to rise, substantial gaps in rheumatic workforce, medical best-practices, and rheumatic care accessibility highlights the need for robust and clinically-applicable evidence regarding visit frequencies to optimize patient outcome, inform policy planning, and project future service needs and healthcare expenditure¹⁰.

With the goal of better understanding visit frequency in rheumatic diseases and how visit frequency might differ across countries, we conducted a systematic literature review to identify studies that have estimated visit frequency (mean number of annual patient visits to rheumatologists and other specialties) for major rheumatic diseases. The review included literature from 1946 to present and systematically summarized and analyzed existing practices of clinical visit frequency with country/state- and disease-specific resolution. Where possible, we created summary statistics but heterogeneity in the included literature precluded a formal meta-analysis.

METHOD

Study Design and Search Strategy

A pre-defined study protocol (https://www.crd.york.ac.uk/PROSPERO/display_record.php?RecordID=306299) was developed and deposited with PROSPERO. Most systematic literature reviews begin with developing a PICO (population, intervention, comparison, outcomes) table. However, we demanded no intervention or comparison, thus a PICO table was unnecessary.

The search strategy included all primary articles with the exception of systematic reviews (including randomized trials, cohort studies, case control studies and cross-sectional studies) that report a measure of clinical visit frequency in rheumatology or general practices. To be as inclusive as possible, no setting or type of rheumatological diagnosis was set a priori. All studies published prior to February 2^nd, 2022, the date on which the search was last ran, were considered eligible.

Due to the highly challenging nature of precisely defining concepts of clinical visit in the literature, multiple iterations of search terms were generated, with the help of Yale MeSH analyzer¹¹ and calibrated with a suite of 11 key papers deemed by the authors to adequately capture the breath of this field. Using the search terms generated from this strategy, one author (YJ) searched the following databases: OVID MEDLINE, CINAHL and WHO Global Index Medicus since their inception with the input of a medical librarian. Please see appendix for the final list of search terms; in brief, it includes the following concepts: house calls, home visits, office visits, clinic visits, frequency (weekly, monthly, yearly), and rheumatology.

Study selection and data extraction

Two reviewers (YJ and DHS) independently scanned all abstracts and potentially eligible full text articles to determine eligibility for inclusion. Discrepancies in judgements between the reviewers were discussed until consensus was reached. Using a standard data extraction sheet, information was extracted on study and patient characteristics including: the year of publication, study period start- and end dates, sources of data (i.e., clinical record, administrative database), geographic location and setting of rheumatology care, rheumatic disease diagnosis, and provider type. The outcome of interest was clinical visit frequency in mean number of visits per patient per year. For a small group of studies, the outcome data were not presented in detail and authors were contacted to help with relevant numbers.

The validity of individual trials was assessed using the risk of bias 2.0 instrument, endorsed by the Cochrane Collaboration, and risk of bias for observational studies was assessed using a modified version of the Newcastle-Ottawa tool (Supplementary Figure 2 and supplementary table 1, respectively). It was modified to include only the outcome assessment measures; as no comparison group was required, the remainder of this tool was not applicable.

Statistical analyses

We categorized studies by rheumatic disease and by country. The preponderance of studies were from the US; we separately categorized studies from the US and studies from non-US settings. Annual visit frequencies from each study were combined using a weighted mean, weighted on the number of patients included. Some studies gave standard deviations or medians and interquartile ranges but not all; thus, standard errors could not be estimated across studies.

RESULTS

Study Eligibility and Selection

As shown in Supplemental Figure 1, a total of 273 records were identified (7 were duplicates), out of which 267 were examined for title and abstract eligibility. Of those, 65 were retrieved for full text screening and eligibility assessment. Out of the 65 papers, 22 did not contain disease-specific visit frequency data, 1 was only available as an abstract, and 14 papers could not be confirmed of their availability of potential unpublished data to calculate outcomes of interest due to inability to make contact despite our inquiries. In the end, 28 studies were included for data extraction.

Study Characteristics

The 28 included studies for clinical visit frequency were equally divided by countries of origin into US and non-US, with the majority (85%) from North America and Europe (see Tables 1 and 2). The studies were published between 1985 and 2021. Rheumatoid arthritis (RA) was the most frequently reported condition (16 paper mentions), followed by systemic lupus erythematosus (SLE) (5 paper mentions) and fibromyalgia (FM) (4 paper mentions). The mean study period was 4.5 years, with a maximum of 29 years and a minimum of less than one year. In 21 (75.0%) of studies clinical visit frequency data were derived from clinic/hospital records, 5 (17.9%) from public healthcare registry, and the remainder 2 (7.1%) from private insurance claims. Half of included studies reported data from a community rheumatology practice setting, 7 did not specify the setting, 6 reported hospital-based outpatient practices, and the remaining one study reported “primary and secondary care” in the Swedish context (which is not easily convertible into the above categories). 74.0% of all studies reported visit frequency data for rheumatologists only, with the remainder including a mixture of general practitioners (GP), internists, and rheumatologists. No noticeable difference in study characteristics were observed between those focusing on RA and the non-RA studies, and the US and non-US studies.

Table 1.

Characteristics of included studies with a reported diagnosis that included but was not limited to rheumatoid arthritis, grouped by location, and types of provider

Author (Year)	Study Period	Data Source	Setting and Location	Provider
US Studies
Yelin (1985)25	1982 – 1983	Clinic record?	Community clinics, Northern California	Rheumatologists-internist
Yelin (1996)26	1982 – 1989	Clinic record	Community clinics, Northern California	Rheumatologists
Criswell (1997)27	1982 – 1983	UCSF Longitudinal RA panel	Community clinics; Northern California	Rheumatologists
Ward (1997)6	1979 – 1981	Stanford Outcomes in Rheumatoid Arthritis study	Community clinic; California	Rheumatologists
Katz (1998)*28	1993 – 1993	Medicare claims (excluding HMOs)	Community clinic, Colorado, Massachusetts, and Virginia	Rheumatologists
Gabriel (2001)29	1987 – 1994	Rochester Epidemiology Project	Hospital-based Outpatient clinic, Minnesota	Rheumatologists
Ethgen (2002)#30	1996 – 1998	Hospital record	Hospital-based Outpatient clinic, Kansas	Rheumatologists
Bartels (2011)31	2004 – 2006	Medicare claims	Rheumatologist visits, any location, random US national sample	Rheumatologists
Accortt (2017)32	2010 – 2013	Truven Health Analytics MarketScan® Database	Rheumatologist visits, any location, US national sample	N/A
Non-US Studies
Chan (2008)33	2001 – 2006	Rheumatology Service database	Outpatient clinic, New Zealand	Rheumatologists
Hagel (2013)34	2001 – 2011	Skåne Healthcare Register	All settings, Sweden	“Physician”
Bengtson (2016)35	2006 – 2010	Hospital record	Primary Care, Sweden	PCP & Secondary Care
McBain (2016)§36	Not available	Hospital record	Hospital-based Outpatient clinic, London, UK	Rheumatologists
Barnabe (2017)†37	1993 – 2011	Public claim database (AHCIP)	Outpatient clinics, Canada	PCP
Kim (2020)38	2019	Clinical record data	Hospital-based Outpatient clinic, South Korea	Rheumatologists
Muskens (2021)39	2014 – 2019	Hospital record	Hospital-based Outpatient clinic, the Netherlands	Rheumatologists
Poggenborg (2021)7	2015 – 2017	Hospital record	Hospital-based Outpatient clinic, Denmark	Rheumatologists

^*OA (any site), Mechanical spinal disorders, Bursitis, tendinitis, FM, PMR, SLE, All rheumatologic diagnoses;

^#OA;

^§PsA;

^†AS, PsD, Crystal-related arthritis;

^?Clinical records obtained via telephone survey

AHCIP = Alberta Health Care Insurance Plan; AS = Ankylosing Spondylitis; PCP = Primary Care Provider; PsD = Psoriatic Diseases; OA = Osteoarthritis; PMR = Polymyalgia Rheumatica; RA = Rheumatoid Arthritis; SLE = Systemic Lupus Erythematosus;

Table 2.

Characteristics of included studies with a reported diagnosis that did not include rheumatoid arthritis, grouped by location, type of provider

Author (Year)	Study Period	Diagnosis	Data Source	Setting and Location	Provider
US Studies
Kremers (2005)40	1970 – 1999	PMR	Truven Health Analytics MarketScan® Database	All settings, Olmsted County, Minnesota	Rheumatologists; Generalists
Molina (2008)41	2003 – 2003	SLE	Triple-S, Inc. insurance record	All settings, Puerto Rico	Rheumatologists
Julian (2009)42	2002 – 2004	SLE	UCSF Lupus Outcomes Study	All settings, US national sample	Rheumatologists
Singh (2011)43	Not available	Crystal-related arthritis	Clinic & hospital record	Outpatient clinic, Hospital outpatient, and VA clinics, San Diego, Cincinnati, and Minneapolis	Rheumatologists
Chandran (2012)44	2018 – 2019	FM	Clinic record (convenience sample)	Outpatient clinic, US national sample	GP, rheumatologists, neurologists or psychiatries
Non-US Studies
Badley (2015)45	2007 – 2008	Inflammatory arthritis	Ontario Health Insurance Plan	N/A, Canada	Rheumatologists
Elek (2015)46	2008 – 2012	General rheum	GYEMSZI Database	Outpatient clinics, Hungary	All physicians; rheumatologists
Andrés (2016)47	2009 – 2010	SpA	Hospital record	Outpatient clinics, Spain	Rheumatologists; Internists
Valent (2020)48	2009 – 2020	GCA	Hospital record	Outpatient clinics, Italy	Rheumatologists; internists
Gendelman (2021)12	1998 – 2014	FM	HMO group record	Outpatient clinics, Israel	Rheumatologists
Winkelmann (2011)13	2011 – 2011	FM	Clinic record	Outpatient clinics, France and Germany	GP & rheum

Abbreviations:

FM = Fibromyalgia rheumatica; GCA = Giant Cell Arthritis; GP = General Practitioner; GYEMSZI = National Institute for the Quality and Organizational Development in Healthcare and Medicines (of Hungary); PMR = Polymyalgia Rheumatica; RA = Rheumatoid Arthritis; SLE = Systemic Lupus Erythematosus; SpA = Spondyloarthropathy.

Clinical Visit Frequencies Across the Disease Spectrum

A wide range of mean clinical visit frequencies were reported across rheumatic diseases and geographic locations. For RA (Table 3), the highest weighted mean visit frequency was reported for US-based rheumatologists at a rate of 5.25 visits per patient year (this unit will apply hereafter unless stated otherwise), followed by US-based non-rheumatologists (4.80), non-US rheumatologists (3.29) and non-US non-rheumatologist (2.74). A temporal decrease in visit frequencies for RA can generally be seen among US-based rheumatologists, from 13.1 in 1985 (at fee-for-service practices) to lowest approximately 2.0 in the first decade of the twenty first century. On the contrary, the reverse trend is observed in non-US-based rheumatologists, from 1.9 in New Zealand in 2008 to 3.4 in South Korea in 2022, to close to 4.0 in Western- and Northern Europe in 2021. The highest frequency of RA clinical visits was reported from a 1985 US study, with 13.1, whereas the lowest was reported in 2001 from a hospital-based outpatient study in Minnesota at a rate of 1.3.

Table 3.

Visit frequencies for rheumatoid arthritis

Author (Year)	Sample Size, patients	Mean Annual Visit Frequency
US Studies - Rheumatologists
Yelin (1985)	N = 626	Fee for service: 13.16
	N = 186	Health Maintenance Organizations: 10.15
Yelin (1996)	N = 798	Fee for service: 9.91
	N = 227	Health Maintenance Organizations: 9.83
Criswell (1997)	N = 310	6.56
Ward (1997)	N = 127	7.20
Katz (1998)	N = 8,027	5.40
Gabriel (2001)	N = 249	1.18
Ethgen, (2002)	N = 642	1.97
Bartels (2011)#	N = 3298	2.40
	Total N = 14,490	Weighted Mean = 5.25
US Studies – Non-Rheumatologists
Accortt (2017)*	N = 6737	4.8
Non-US Studies -- Rheumatologists
Chan (2008)	N = 26	New RA: 2.60
	N = 177	Flaring with new DMARD: 2.55
	N = 170	Stable on bDMARD: 1.48
	N = 75	Stable on NSAID or csDMARD: 1.22
Kim (2020)	N = 378	3.40
Muskens (2021)§	N = 1059	3.8
Poggenborg (2021)	N = 117	3.5
	Total N = 2002	Weighted Mean = 3.29
Non-US-based Non-Rheumatologists
Barnabe (2017)	N = 93490	1.00
Hagel (2013)	N = 3977	8.3
Bengtsson (2016)	N = 7712	Primary Care Provider: 8.0
		Secondary Care: 13
	Total = 105179	Weighted Mean = 2.74

^#Represents the total visits for any reason to a rheumatologist among patients meeting RA definitions using ICD algorithms

^*Physician speciality not available

^§Interrupted time series; Calculated using baseline data prior to intervention.

Abbreviations:

bDMARD = biologic DMARD; csDMARD = conventional syntheticDMARD; DMARD = Disease-modifying antirheumatic drug;

For FM (Table 4), the weighted mean visit frequency for US-based rheumatologists was 1.8 was substantially higher than that of the non-US-based rheumatologists, 0.40. However, if analyzed with the exclusion of the Israeli study¹², Western European visit frequencies were higher than that of the US. A 2012 study using a large US national sample demonstrated a positive association between visit frequency and FM severity measured with the Fibromyalgia Impact Questionnaire score: patients with mild FM met with their rheumatologists a mean of 2.6 times per year, whereas those with severe FM had 6.6 visits per year – the highest among all included FM studies. This trend is replicated in Germany and France in the 2011 European study¹³. On the opposite end of the spectrum, the 2021 Israeli study reported the lowest FM annual visit frequency at 0.32.

Table 4.

Visit frequencies for fibromyalgia

Author (Year)	Sample Size	Mean Annual Visit Frequency (# per patient years)
US Studies, rheumatologists
Katz (1998)	N = 2971	1.8
US Studies, non-rheumatologists
Chandran (2012)*;§	N = 21	Mild FM: 2.7
	N = 49	Moderate FM: 5.2
	N = 133	Severe FM: 6.9
Non-US Studies, rheumatologists
Gendelman (2021)	N = 14269	0.32
Winkelmann (2011)	N = 88	France Total: 2.9
	N = 17	Mild: 3.0
	N = 33	Moderate: 2.5
	N = 38	Severe: 3.4
	N = 211	Germany Total: 4.9
	N = 52	Mild: 4.1
	N = 66	Moderate: 4.6
	N = 93	Severe: 5.5
	Total = 14568	Weighted Mean = 0.401

^*Severity-level classification was based on Fibromyalgia Impact Questionnaire score: 0 to less than 39 = mild; 39 to less than 59 = moderate; 59 to 100 = severe;

^§:primary care physicians, rheumatologists, neurologists, and psychiatrists

For SLE (Table 5), no non-US studies met the eligibility criteria. There was a dramatic difference between the visit frequencies to non-rheumatologists (12.3) and rheumatologists (3.2). Compared to medication adherent patients, non-adherent patients were reported to more frequently visit rheumatologists (4.34 vs 3.24) and other physicians (22.43 vs 18.52), and less frequently visit GPs (6.27 vs 6.43); adherent and non-adherent patients were defined based on reporting that forgetting medications was never a problem versus reporting that it was a problem at least some of the time. The lowest among all reported SLE annual visit frequencies was from the Commonwealth of Puerto Rico, at 1.5 visits per patient year.

Table 5.

Visit Frequencies to rheumatologists for systemic lupus erythematosus

Author (Year)	Sample Size	Mean Annual Visit Frequency (# per patient years)
US Rheumatologists
Katz (1998)	N = 783	2.7
Julian (2009)	N = 454	3.24 (adherent)
	N = 380	4.34 (non-adherent)
Molina (2008)	NO DATA	3.0
	Total = 1617	Weighted Mean = 3.24
US Non-rheumatologists
Julian (2009)	N = 454	3.80 (adherent, GP visits)
	N = 454	18.52 (adherent, other health care provider)
	N = 380	4.78 (nonadherent, GP visits)
	N = 380	22.43 (nonadherent, other health care provider)
Molina (2008)	NO DATA	1.5
	Total = 1668	Weighted Mean = 12.3

Due to the paucity of studies pertaining to each individual condition, all non-RA arthritis were grouped into Table 6. Only two US studies reported on clinical visit frequencies of OA, with a weighted average of 2.15. Striking differences can be seen between the rates of clinical visits between US regions and Alberta, Canada for crystal-related arthritis, with the former reporting 2.0 visits to rheumatologists per patient year while the latter reporting 0.034 visits per patient year, including both rheumatologists and internists. Visit frequencies to rheumatologists and non-rheumatologists were similar for psoriatic diseases (PsD), around 0.25 per patient per year. This is not the case for spondyloarthropathies (SpA), which showed a four-fold difference between rheumatologist visits and non-rheumatologists.

Table 6.

Visit Frequencies for other arthritis

Author (Year)	Sample Size	Visit Frequency (# per patient years)	Diagnosis	Comments
US Studies
Singh (2011)	N = 285	1.85	Gout
Katz (1998)	N = 15715	2.2	OA	OA for any site; median = 2
	N = 7334	2.1	Mechanical spinal disorders	Median = 1
		1.9	Bursitis, tendinitis	Median = 1
		3.3	PMR	Median = 2
		3.8	All rheumatologic diagnoses	Median = 3
Ethgen, (2002)	N = 395	0.097	OA
Kremers (2005)	N = 364	0.84	PMR	Rheumatologist
		3.8	PMR	generalist
Elek (2015)	N = 430000	0.021	general rheum	extensive margin Fixed-effects logit equation
		0.0068	general rheum	intensive margin Fixed-effects truncated Poisson equation
		0.19	general rheum	Pooled zero-inflated negative binomial model
Valent (2020)	N = 208	0.71	GCA	Rheumatologist
		0.65	GCA	Internal Medicine
Non-US Studies
McBain (2016)	N = 48	2.15	PsA mixed w/RA
Barnabe (2017)	N = 6040	0.27	Psoriasis & psoriatic arthritis	Rheum
	N = 6040	0.25	Psoriasis & psoriatic arthritis	Non-rheum
	N = 44221	0.51	Crystal-related arthritis	PCP; crystal-related arthritis
		0.034	Crystal-related arthritis	rheumatologists or internists; crystal-related arthritis
	N = 7685	0.37	AS	Rheumatologist
		0.22	AS	Non-Rheumatologist
Andrés (2016)	N = 1168	2.0	SpA	rheum
	N = 1168	0.50	SpA	non-rheum
Badley (2015)	NO DATA	0.0069	Inflammatory Arthritis	All physicians
	NO DATA	0.012	All Arthritis	All physicians

DISCUSSION

This systematic review identified and summarized estimated visit frequencies for major rheumatic diseases reported in the literature from 1946 to present. This exercise helps shed light on the sparse data concerning rheumatology visit frequencies. We have attempted to compare visit frequencies across time, across disease categories, and across geographic locations. Of note, our comprehensive results spanned five out of seven World Bank region groups¹⁴ (South America and sub-Saharan Africa sources were absent). Out of the 28 included studies from a total of 272 unique records examined, the great majority were from North America and European centers, involved RA, featured community rheumatology practices, and consisted solely of rheumatologists. Weighted mean visit frequencies for RA was 5.25, 4.80, 3.29 and 2.7 per year for US rheumatologists, US non-rheumatologists, non-US rheumatologists and non-US non-rheumatologists respectively. Fibromyalgia patients made 1.80 and 0.40 annual visits to US and non-US rheumatologists respectively. For US rheumatologists and non-rheumatologists, visit frequency for SLE was 3.2 and 12.3, respectively.

Two general trends observed in our data warrant examination. First, visit frequencies are higher in the US as compared to non-US studies for all conditions where such comparisons are possible, especially for RA and FM (Tables 3 and 4). The US has one of the highest per capita rheumatology workforces at 1.78 per 100,000. The American College of Rheumatology (ACR) workforce report describes a deficit but the US rheumatology workforce appears to have increased from 4,049 US rheumatologists in 2005 data¹⁵ to 5602 in 2022¹⁶. When it comes to per capita rheumatologists, focusing on the countries included in Table 3, the US, with 1.78 rheumatologists per one hundred thousand population¹⁰, exceeds that of New Zealand (0.59)¹⁷, South Korea (0.60)¹⁸, the Netherlands (0.61)¹⁹. Differing guideline recommendations could also play a role in the differing visit frequency data: while both ACR and the European League Against Rheumatism (EULAR) recommend treat-to-target strategies for RA management, the ACR gave a loose, minimum monitoring frequency of at least every 6 month²⁰, whereas EULAR specified an interval of 1 – 3 month during active disease²¹, potentiating variable visit frequencies due to individual provider preferences. Other disease treatment guidelines of ACR and EULAR provide no clear recommendations on visit frequency.

Second, it is evident from Table 3 that a clear trend exists for reduction in rheumatologist visits for RA over time, from 13.1 visits per year in 1985 to 2.4 visits per year in 2011. The introduction of conventional synthetic DMARDs such as methotrexate fundamentally changed RA management and patient outcomes²², and the clinical approval by the FDA of TNF inhibitors in 1999 heralded the biologic DMARD age of RA treatment. Despite the need for intensive monitoring at the early stage of initiation, advanced therapies allow for sustained long-term low disease activity or remission that would gradually reduce the need for frequent rheumatologist visits – which is reflected by the temporal trend in Table 3.

Our systematic review has several strengths. For one, it is the first and only paper we are aware that provides a global, comprehensive picture pertaining to clinical visit frequencies across the spectrum of common rheumatic conditions, covering a consequential four-decade period in the history of rheumatology from 1985 to 2021. An additional strength of this review is the search term engineered using Yale MeSH analyzer and calibrated with key citations. Finally, a third strength of our review beyond its strict adherence to the PRISMA workflow, is its broad coverage of major medical literatures (MEDLINE), allied health literature (CINHAL), and global grey literature (WHO Globus Index Medicus).

Our review has several limitations. Due to the nature and heterogeneity of the clinic visit frequency literature, most if not all of our included papers reported study visit frequency as a secondary outcome. Not being the main focus of their respective reports, visit frequency data tended to lack details such as precise provider types (including advanced practice providers), measurements of statistical variance, disease severity, settings and context such as clinical visits versus laboratory visits, although the latter may diminish in the future with the emergence of novel care models such as at-home blood sampling²³ and one stop clinics²⁴. Future studies in rheumatology visit frequencies should strive to include many if not all of the aforementioned details to allow for a more granular understanding of the factors influencing the manner by which care is being utilized. It is also possible that visit frequency will continue to change as technology enables us to communicate with patients in previously unobserved ways. Furthermore, we were not able to assess reports written in languages other than English. Finally, the lack of existing and validated quality assessment tools for qualitative health service research papers compelled us to adopt and modify the Ottawa-Newcastle scale, which may not be as useful in the context of this review.

In conclusion, we found the clinical visit frequency literature to be sparse and methodologically heterogenous, focusing mainly on RA and published mainly from the US. Studies reported more frequent visits by US vs non-US rheumatologists and universally showed a decreasing temporal trend of visit frequencies in the US. Future studies are strongly encouraged to focus on visit frequencies across time as the primary outcome, provide rigorous definitions for the nature of reported visits, clarify in detail the setting of visits and type of providers, and, for any quantitative data, report on measures of statistical variance. To advance the field of clinical visit frequency research, there is a need to develop validated methods to link patient disease activity scores (or prognostic data) with visit frequencies. Doing so would shed light on the implications of clinic visits on disease progression (or the lack thereof), thereby allowing for determination of the ideal visit frequency for each common rheumatic diseases by professional societies in their guidelines, potentially eliminating unnecessary visits and hence healthcare expenditure, and optimizing resource utilization.

SIGNIFICANCE AND INNOVATION.

• Clinical visits are a fundamental aspect of rheumatic disease care, but recommendations for appropriate visit frequencies are largely absent from guidelines, scarcely studied, and inconsistently reported.

• We included 28 relevant manuscripts in this systematic literature review; the included studies were equally divided between US and non-US and were published between 1985 and 2021.

• For RA, the average annual visit frequencies were 5.25 for US rheumatologists, 4.80 for US non-rheumatologist, 3.29 for non-US rheumatologists, and 2.74 for non-US non-rheumatologists.

• Trends suggest more frequent visits in the US and less frequent visits in recent years.

APPENDIX I: Search Terms

(Home Visits.mp. or exp House Calls OR office visits.mp. or exp Office Visits OR Clinic Visits.mp. or exp Clinic Visits/ OR primary health care.mp.) AND (visit*.mp. and ((frequenc* or “use” or trend* or characteristics or wellness or annual).mp. or *annual/ or monthly.mp. or weekly.mp. or daily.mp. or decline.mp. or increase.mp.) [mp=title, abstract, original title, name of substance word, subject heading word, floating sub-heading word, keyword heading word, organism supplementary concept word, protocol supplementary concept word, rare disease supplementary concept word, unique identifier, synonyms]) and (rheumatology or rheumatoid or rheum*).mp. [mp=title, abstract, original title, name of substance word, subject heading word, floating sub-heading word, keyword heading word, organism supplementary concept word, protocol supplementary concept word, rare disease supplementary concept word, unique identifier, synonyms].