Abstract
Objective To determine if continuity of rheumatology care influences rates of emergency department (ED) visits and hospitalizations in patients with rheumatoid arthritis (RA).
Methods A closed inception cohort of patients with RA diagnosed between 2000 and 2009 were followed until December 31, 2019. During the first 5 years following diagnosis, we categorized patients into 3 rheumatology care continuity groups (high, intermediate, and not retained in rheumatology care). Using a landmark analysis, we compared rates of ED visits and hospitalizations during follow-up. Multivariable Poisson regression models were used to estimate rate ratios (RRs), adjusting for demographics, comorbidities, and health services access and supply measures.
Results The cohort included 38,528 patients, of which 57.7% (n = 22,221) were classified in the high rheumatology continuity group, 17.2% (n = 6636) were in the intermediate group, and 25.1% (n = 9671) were not retained in rheumatology care. Relative to the high continuity group, both the intermediate and nonretention groups had higher ED rates (RR 1.14, 95% CI 1.08-1.20, and RR 1.12, 95% CI 1.08-1.16, respectively). The intermediate group also experienced higher adjusted hospitalization rates (207.4, 95% CI 203.0-211.8 per 1000 person-years [PY]) than the high continuity group (193.5, 95% CI 191.4-195.6 per 1000 PY).
Conclusion Patients with RA with higher continuity of rheumatology care had lower rates of ED visits and hospitalizations compared to those who did not receive continuous rheumatology care during the first 5 years of follow-up. These findings provide evidence to support the value of early and continuous rheumatology care for reducing hospitalizations and ED visits.
- continuity of care
- emergency visits
- health services research
- hospitalizations
- rheumatoid arthritis
- rheumatology
For rheumatoid arthritis (RA), a plethora of evidence has demonstrated the value of early diagnosis and ongoing assessments and treatment by a rheumatologist in terms of improving clinical outcomes.1,2 Clinical studies have shown that patients who persist on therapy have superior clinical outcomes3 and population-based studies have shown that patients under the care of a rheumatologist are more likely to receive treatment and persist on therapy.4,5 Individuals with RA who remain under rheumatology care may also benefit in terms of improved extraarticular disease and comorbidity management. Therefore, enhancing continuity of rheumatology care is likely a beneficial strategy for reducing costly acute care services such as hospitalizations and emergency department (ED) visits.
To date, few studies have evaluated the effects of continuity of rheumatology care among patients with RA. For individuals with other chronic diseases, sustained continuity of care is frequently linked to improved patient outcomes, quality of care, adherence to therapy, receipt of preventive care services, and the appropriate use of other healthcare resources such as specialty, laboratory and diagnostic services, and reductions in acute care services.6,7
Therefore, we assessed if continuity of rheumatology care influences rates of ED visits and hospitalizations in patients with early RA.
METHODS
Study design. This was a retrospective population-based closed inception cohort study. Individuals diagnosed with RA between January 1, 2000, and December 31, 2009, were followed until December 31, 2019. Patients with RA were categorized into continuity of care groups according to their level of care continuity up to their fifth year following diagnosis. A landmark analysis was undertaken to compare hospitalization and ED rates in patients with RA with varying levels of continuity of care, with 5 years after diagnosis as the landmark date (Supplementary Figure S1, available with the online version of this article).
Setting. All Ontario residents are insured under a single-payer healthcare system, which covers all medically necessary health services, including rheumatology consultations, hospitalizations, and ED visits. These contacts for health services are recorded in administrative databases. The use of the data in this study was authorized under section 45 of Ontario’s Personal Health Information Protection Act, which does not require review by a research ethics board.
Data sources. The Ontario Health Insurance Plan (OHIP) Claims Database was used to identify diagnoses associated with physician services. Patient demographic and vital statistic information was ascertained from the Registered Persons Database. Rheumatologists were identified using the validated ICES physician registry. Hospitalizations were identified using the Canadian Institute for Health Information Discharge Abstract Database. ED visits were ascertained from the National Ambulatory Care Reporting System. These datasets were linked using unique encoded identifiers and analyzed at ICES (https://www.ices.on.ca).
Study population. Patients with RA were identified from the Ontario Rheumatoid Arthritis Dataset, a population-based registry assembled using a validated algorithm that has a 78% positive predictive value (PPV), 78% sensitivity, and ~100% specificity at identifying RA.8,9 All patients were required to have at least 3 physician claims for RA over 2 years with at least 1 by a rheumatologist.
We included incident patients with RA aged ≥ 18 years at the time of their first RA code between January 1, 2000, and December 31, 2009. Patients were excluded if they had missing demographic information, were a non-Ontario resident at the date of their first RA code, did not have OHIP eligibility in the 5-year period prior to their first RA code (to exclude potential prevalent RA patients moving to Ontario), or if they died or outmigrated prior to their landmark date.
Continuity of rheumatology care definition. Continuity of rheumatology care was measured from the first RA code up to the start of the fifth year (landmark date) following diagnosis, representing 4 annualized measurement periods (Supplementary Table S1, available with the online version of this article). Patients were classified into 1 of 3 categories, determined a priori: (1) high continuity, (2) intermediate continuity, and (3) not retained in rheumatology care. High continuity was defined as patients with at least 2 rheumatology consultations during the first year following diagnosis, followed by at least an annual visit in the subsequent measurement periods/years. Intermediate continuity was defined as having an annual rheumatology visit on > 50% of the measurement periods. Patients were classified as not retained if they failed to meet the intermediate continuity definition. The high continuity definition was informed by clinical guidelines and health system performance measures, recommending that patients with RA require, at a minimum, an annual rheumatology consultation regardless of disease activity, and more frequent monitoring following new treatment initiations (which is the case in the first year following diagnosis).4,10,11
Outcomes. All-cause unscheduled hospitalizations and ED encounters were ascertained from the landmark date (day 1460) onward.
Statistical analysis. We conducted descriptive analyses to summarize and compare the clinical characteristics of the patient cohort at the landmark date, according to the level of care continuity. Analyses were performed to separately determine ED visit rates and unscheduled hospitalization rates determined from the landmark date to the end of follow-up for each group. Patients were followed from the landmark date until death, outmigration, or until the end of study period (December 31, 2019). Person-years (PY) of follow-up time were accumulated and used as the denominator for the rates of the outcomes. Crude rates for ED visits and hospitalizations were measured by dividing the total occurrence by the cumulative PY of follow-up time, stratified by level of care continuity. Age- and sex-adjusted rates were also measured for each outcome.
To estimate rate ratios (RRs) across care continuity groups, modified Poisson regression models with robust error variances were used to adjust for patient demographics, comorbidities, having a primary care physician, and rheumatology access and supply measures. Patient sociodemographic covariates included in each model were age, sex, living in a rural area (based on each patient’s postal code and a community population size < 10,000 residents) vs urban area, and income quintiles (defined as the patient’s neighborhood median household income quintile from the Statistics Canada Census). The Johns Hopkins ACG® System was used to identify the number of Aggregated Diagnosis Groups to characterize the morbidity of an individual based on hospitalizations, ED visits, and physician claims data. Additional specific comorbidities included the presence of hypertension, diabetes, coronary artery disease, cardiovascular events, chronic obstructive pulmonary disease or asthma, kidney disease, cancer, and extraarticular manifestations of RA. As a proxy for RA severity, presence of extraarticular RA manifestations (lung, hematological, cardiac, eye, or dermatological complications, entrapment syndromes, neuropathies, or amyloidosis) was based on having at least 1 hospital or 2 diagnosis claims in the 3 years prior. Rheumatology access and/or supply variables included whether patients lived at remote distances from a rheumatologist (> 100 km) and the regional supply of rheumatologists (optimal vs suboptimal), with optimal regional rheumatology supply defined as residing in a region with at least 1 rheumatologist per 75,000 adults. Last, a binary measure of whether the patient had a designated family physician and prior joint replacements were controlled for. Detailed definitions of the covariates are further detailed in Supplementary Table S2 (available with the online version of this article). Adjusted RRs were separately determined for each outcome, using the high continuity group as the referent group. All statistical analyses were performed using SAS version 9.3 (SAS Institute).
RESULTS
Descriptive and clinical characteristics. From a total of 44,822 individuals with a diagnosis of RA between 2000 until 2009, 38,528 patients were retained for this analysis (Supplementary Figure S2, available with the online version of this article). Among the 38,528 patients with RA, 22,221 (57.7%) individuals were classified as having high continuity of rheumatology care, 6636 (17.2%) individuals received intermediate continuity of rheumatology care, and 9671 (25.1%) individuals were classified as nonretained.
Table 1 compares the clinical characteristics of patients categorized by continuity exposure. Relative to the other groups, the high continuity group had more females (72.8%), were older (mean 59.7 yrs), resided in urban areas (87%) with higher income quintiles, and had better rheumatology supply. Patients within the intermediate group were most likely from rural areas (18.2%), resided farther away from rheumatologists, and had the highest global burden of comorbidity. The nonretained group had the highest percentage of males (33.6%), were more likely to reside in regions with suboptimal regional rheumatology supply, and were less likely to have a family physician. This group also had the lowest burden of comorbidities.
Clinical characteristics of the cohort at landmark date.
ED visits. The age- and sex-adjusted ED visit rate was lowest for the high continuity group (635.7/1000 PY, 95% CI 630.7-640.7), followed by the nonretention group (699.7/1000 PY, 95% CI 690.6-708.8; Table 2). Patients in the intermediate continuity group had the highest ED visit rate (773.7/1000 PY, 95% CI 759.8-787.7).
Emergency department (ED) rates and rate ratios (RRs) following landmark date stratified by continuity of care groups.
The ED visit RRs from the univariate model were 1.09 (95% CI 1.05-1.13) for the nonretention group, and 1.21 (95% CI 1.14-1.28) for the intermediate continuity group, referent to the high continuity group. Relative to the high continuity group, ED RRs from the multivariate model controlling for all covariates were 1.12 (95% CI 1.08-1.16) for the nonretention group and 1.14 (95% CI 1.08-1.20) for the intermediate continuity group.
Hospitalizations. The age- and sex-adjusted hospitalization rates/1000 PY were highest for patients with intermediate continuity (207.4, 95% CI 203.0-211.8) compared to 193.5 (95% CI 191.4-195.6) for those with high continuity, and 180.8 (95% CI 177.5-184.1) for those not retained in rheumatology care (Table 3). Relative to the high continuity group, the hospitalization RRs for the multivariate model controlling for all covariates were 0.94 (95% CI 0.91-0.98) for the nonretention group and 1.02 (95% CI 0.98-1.06) for the intermediate continuity group.
Hospitalization rates and RRs following landmark date stratified by continuity of care groups.
DISCUSSION
This study examined whether continuity of rheumatology care in early RA influences subsequent hospitalizations and ED visits. Our results provide suggestive evidence that higher rheumatology care continuity results in lower rates of ED visits and hospitalizations, compared to those who did not receive early contact with a rheumatologist and those who had fewer or interrupted rheumatology visits during follow-up (the intermediate group).
There are several possible underlying mechanisms as to why individuals with high rheumatology continuity had fewer ED visits and hospitalizations, relative to those with fewer visits (the intermediate group). First, individuals with early and high rheumatology care continuity are most likely to receive RA care according to recommended best practices and are therefore most likely being treated within the early onset of disease. Patients with RA who are treated early are most likely to achieve the best short- and long-term clinical outcomes, as early therapy has the capability of altering the disease process and influence its disease progression.1 Further, patients with RA require frequent reassessment of disease activity by a rheumatologist and adjustment of therapy, ensuring tolerance and adherence. Individuals who lack continuous rheumatology care are less likely to have disease-modifying antirheumatic drug (DMARD) dispensations,4 and discontinuation of therapy can be associated with high relapse rates.12,13 Thus, higher rheumatology continuity can have a protective effect for patients with RA as their disease course will be optimally managed. Patients with RA who have established high care continuity with their rheumatologist may seek care for their disease flares, or assistance with other medical needs, directly with their rheumatologist rather than through the ED. Based on our results, it remains unclear whether subgroups of patients within the high continuity group can exhibit further improvements in outcomes and is an area of further research (for instance, additional benefits for visits every 6 months as opposed to annual visits).
Interestingly, patients with RA within the nonretention group experienced lower hospitalization rates but higher ED visit rates, relative to the high continuity group. The lower rates of hospitalizations in this group may reflect those with milder RA disease phenotypes, those with incorrect diagnoses, or those with self-limiting forms of arthritis. Previous studies have suggested that as many as 15% to 25% of patients with early RA experience sustained drug-free remission periods,12,13 and as many as 42% to 82% of patients may achieve remission after 6 months of treatment.14 It may also be possible that patients in the nonretention group are more likely to be misclassified as RA. Based on previous validation research, the PPV of the RA case definition used in this study was estimated to be 78% based on 2 independent validation studies/samples, meaning that 22% of patients could be false positive RA cases.8,9 According to the validation study, misclassified patients most commonly had query RA diagnoses, arthralgia currently being monitored for the development of RA, or other forms of inflammatory arthritis.8 However, the increased ED visits in the nonretention group may also signal that the ambulatory care needs for this subset of the population are not being met. Indeed, individuals who were not retained in rheumatology care were more likely to lack a primary care physician, be male, and reside in regions with suboptimal regional supply of rheumatologists. Therefore, these individuals appear to have less access to care, or are receiving less care, and their increased ED visit utilization may be a reflection that these individuals are receiving fewer preventive health services. In a previous study, we have found that increasing patient age, male sex, lower socioeconomic status, higher comorbidity burden, and having an older rheumatologist were factors independently associated with not being retained in rheumatology care.4 Targeted efforts are needed to reduce the disparities that these subpopulations are experiencing.15-17
To our knowledge, this is one of the first studies to explore how continuity of rheumatology care affects patient outcomes in RA. However, the results of this study can be supported by studies evaluating treatment adherence and RA outcomes. Although not completely synonymous, continuity of care may be considered a proxy for treatment adherence. Studies have supported the association between higher treatment adherence with lower rates of hospitalizations, ED visits, and healthcare costs.18-20
Strengths of this study include the use of population-based health data in the setting of a publicly funded health system for the entire population, with long and comprehensive follow-up. The results of this study must be considered in light of several limitations. First, despite using a validated case definition, misclassification of patients with RA is possible. Second, 2 potential clinical confounders that we were unable to adjust for (due to lack of data availability) include disease activity and treatment. It is possible that patients with less active or less severe RA are less likely to sustain ongoing rheumatology care than patients with more severe disease. Another potential confounder or effect modifier is RA treatment. As pharmacy claims data are only available on a subset of patients (age > 65 yrs) in Ontario, we were unable to control for DMARD use (or other therapies). There are also possible differences between rheumatology care in the specified study period of 2000 to 2009 compared to current care. The treat-to-target recommendations that include early and targeted treatment for RA were established in the early 2010s, and thus contemporary patients with early RA may differ in outcomes.21 Future research is also needed to evaluate other dimensions of continuity of care (and alternative definitions to optimally define continuity) on outcomes for patients with RA.
In conclusion, patients with incident RA who received early and more frequent rheumatologist visits in the first year of diagnosis, followed by at least an annual rheumatology visit during the first 5 years following diagnosis, were subsequently found to have fewer hospitalizations and ED visits compared to patients who had inconsistent rheumatology care. This study illustrates that when rheumatology care is delayed, disrupted, or is less frequent, patients with RA are at greater risk for ED visits and hospitalizations.
ACKNOWLEDGMENT
The study was supported by ICES (formerly known as the Institute for Clinical Evaluative Sciences, www.ices.on.ca), which is funded by an annual grant from the Ontario Ministry of Health (MOH) and the Ministry of Long-Term Care (MLTC). This document used data adapted from the Statistics Canada Postal CodeOM Conversion File, which is based on data licensed from Canada Post Corporation, and/or data adapted from the Ontario MOH Postal Code Conversion File, which contains data copied under license from Canada Post Corporation and Statistics Canada. Parts of this material are based on data and information compiled and provided by the MOH, MLTC, and the Canadian Institute for Health Information. The analyses, conclusions, opinions, and statements expressed herein are solely those of the authors and do not reflect those of the funding or data sources; no endorsement is intended or should be inferred.
Footnotes
This study was funded by an operating grant from the Canadian Initiative for Outcomes in Rheumatology Care (CIORA), who played no role in the design or conduct of the study, other than providing peer review of the study proposal. LE receives support from an Early Researcher Award from the Ontario Ministry of Research, Innovation and Science and is Canada Research Chair (Tier 2) in Inflammatory Rheumatic Diseases. CEHB receives support from the Arthritis Society Stars Career Development Award funded by the Canadian Institutes of Health Research Institute of Musculoskeletal Health (STAR-19-0611/CIHR SI2-169745). JW receives support from the Arthritis Society Stars Career Development Award (STAR-19-0610).
LE received educational and research grants and consultation fees from AbbVie, UCB, Eli Lilly, Novartis, Sandoz, and Pfizer (unrelated to this work). JCT received educational and research grants and consultation fees from AbbVie/Abbott, Amgen, CaREBiodam, Celgene, Centocor, Eli Lilly, Janssen, Medxus/Medac, Merck/MSD, Novartis, Pfizer, and Sanofi. The remaining authors declare no conflicts of interest relevant to this article.
- Accepted for publication January 6, 2023.
- Copyright © 2023 by the Journal of Rheumatology