key: cord-0920626-iwd0l05p authors: Kao, Yu-Hsiang; Tseng, Tung-Sung; Ng, Yee-Yung; Wu, Shiao-Chi title: Association between continuity of care and emergency department visits and hospitalization in senior adults with asthma-COPD overlap date: 2018-11-15 journal: Health Policy DOI: 10.1016/j.healthpol.2018.11.005 sha: 02612805eaebdc70fd7dc11ff6e091c0ca9dfdac doc_id: 920626 cord_uid: iwd0l05p OBJECTIVE: To investigate associations between continuity of care (COC) and emergency department (ED) visits and hospitalization for chronic obstructive pulmonary disease (COPD) or asthma among elderly adults with asthma-COPD overlap (ACO). METHODS: A retrospective cohort study was performed using the Taiwan National Health Insurance research database. A total of 1141 ACO patients aged ≥65 years during 2005–2011 were observed and followed for 2 years. The Bice and Boxerman COC index (COCI) was used to evaluate COC by considering ambulatory care visits duo to COPD or asthma in the first year; ED visits and hospitalization for COPD or asthma were identified in the subsequent year, respectively. The COCI was divided into three levels (COCI < 0.3= low, 0.3 ≤ COCI<1=medium, COCI = 1=high). The Cox model was used to estimate the hazard ratio (HR) for ED visits and hospital admissions due to COPD or asthma. RESULTS: The average COCI was 0.55. 21.3% patients received outpatient care from a single physician. Compared to patients with high COC, those with low and medium COC had a higher risk of ED visits (aHR = 2.80 and 2.69, P < .01) and admissions (aHR = 1.80 and 1.72, P < .05). CONCLUSION: Increasing COC is beneficial for elderly patients with ACO in disease management. Policymakers could create effective pay-for-performance programs for the elderly ACO population to enhance COC and improve care outcomes. Asthma-chronic obstructive pulmonary disease (COPD) overlap (ACO) is characterized by persistent airflow limitation with several similar features associated with asthma and COPD [1] . Global Initiative for Asthma (GINA) and Global Initiative for Chronic Obstructive Lung Disease (GOLD) recognized ACO to notice clinicians, researchers and regulators being aware of these patients because of complex conditions [2] . Studies of ACO reported the prevalence ranges from 0.9% to 11.1% among general population around the world [3] . The disease burden and healthcare utilization such as emergency department (ED) visits and hospital admissions for patients with ACO has been found to be higher than those with asthma or COPD alone [4] [5] [6] [7] [8] [9] . While the importance of caring for patients with ACO has become a critical issue, there is limited information regarding treatment in ACO patients to reduce the resulting healthcare utilization. Additionally, there was a greater increase in the proportion of patients with ACO with increasing age [7, 8, 10] . Hence, an investigation of elderly patients with ACO is necessary because our society is aging. Previous literature demonstrated that appropriate asthma or COPD control leads to high quality of life and lower healthcare utilization [11] . Therefore, the Agency of Healthcare Research and Quality (AHRQ) [12] and the Centers for Medicare and Medicaid Services (CMS) [13] have identified admissions for COPD or asthma in older adults as a prevention quality indicator (PQI), which may be avoided under more effective treatment in outpatient care. In regard with disease management especially for chronic diseases, continuity of care (COC), is a crucial element of primary care [14] to improve clinical outcomes by consistent and seamless treatment between patients and physicians over time [15] . Previous studies revealed COC is associated with healthcare outcomes in elderly patients with asthma or COPD, respecitively [16] [17] [18] [19] [20] [21] [22] [23] . These studies supported higher COC is associated with fewer ED visits [16] [17] [18] and hospitalizations [16, [18-23] , and lower healthcare cost [16, 18] . However, the potential impact of COC and the therapeutic relationship between physician and patient, in which https://doi.org/10.1016/j.healthpol.2018. 11 .005 0168-8510/© 2018 Elsevier B.V. All rights reserved. improvements in healthcare outcomes such as ED visits or hospital admissions due to asthma or COPD for elderly adults with ACO, is not sufficiently understood. Therefore, to examine the association between COC and care outcomes among ACO population is worthwhile for patients, clinicians and policy regulators in disease management and policy development. The Taiwanese government implemented a single-payer National Health Insurance (NHI) program in March 1995 and approximately 99% of Taiwan's 23 million residents are enrolled in this program [24] . The NHI program provides universal and compulsory health insurance to strengthen accessibility of health care services and to reduce patients' financial barriers. Under the NHI program, patients can freely visit any health care provider at any level without referral [25] . Due to features of this healthcare system, patients with ACO may seek out one or more physicians to care for their health, which leads increase outpatient care visits. However, there is no evidence reporting the relationship between COC and healthcare outcomes for these elderly patients. Therefore, exploration of the COC and its effects on healthcare outcomes among elderly ACO patients may offer new contributions to extend current literature regarding healthcare outcomes. For this reason, the aim of this study is to investigate whether high COC reduces the risk of ED visits and hospital admissions for COPD or asthma among elderly patients with ACO under a national insurance system. This retrospective cohort study is based on claim data collected from 2004 to 2013 from the longitudinal Health Insurance Database for 2010 (LHID2010) and consists of a million randomized beneficiaries in 2010 from the entire National Health Insurance (NHI) enrollee maintained by the National Health Research Institute. All selected beneficiaries were retrospected to 1996 and followed to 2013. The claim data contain comprehensive personal and healthcare care records, such as cryptographic identification for each patient and physician, patient's gender, date of birth, and International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) codes for primary and secondary diseases. There were no significant differences in the distributions of age or gender between patients in the LHID2010 and the nationwide databases [26] . Therefore, the LHID2010 database is representative of the entire population. The years between 2004 and 2013 were selected in this study because of Severe Acute Respiratory Syndrome (SARS) outbreak in Taiwan from March 14 to May 22 in 2003 [27] . The institutional review board at National Yang-Ming University of Taiwan approved this study (IRB Approval Number: YM103047E). There were 15,762 patients aged ≥65 years who either had at least 2 outpatient visits or 1 inpatient admission for COPD as a primary diagnosis (ICD-9-CM codes: 491.xx, 492.xx, 496.xx) [5, 6, 28, 29] identified during the period from January 1, 2005 , to December 31, 2011. The index date was set as the earliest date for patients who had a record during this period. Patients who were not continuously enrolled in the NHI program (n = 954) until the end of study period were excluded. To identify ACO patients, patients without ≥2 outpatient visit claims or 1 inpatient admission for asthma (ICD-9-CM code 493.xx) in a prior year of index date [29, 30] were excluded (n = 13,409). In order to calculate a stable COC value that requires ≥3 outpatient visits; therefore, patients had <3 outpatient visits during the COC period were excluded (n = 222). In total, 1141 patients were included (Fig. 1 ). To avoid time-dependent bias and incorrect conclusions [31] , we followed each patient for 2 years after the index date. The first year was the COC period, and the subsequent year was the outcome period ( Fig. 2 ). Two interesting outcomes in our study concerned ED visits and hospital admissions for either COPD or asthma. According to the PQI for COPD or asthma in older adults proposed by AHRQ and CMS, a certain number of ED visits and hospital admissions for asthma (ICD-9-CM codes: 493.xx in the primary diagnosis) or COPD ((1) ICD-9-CM codes: 4910, 4911, 49120, 49121, 4918, 4919, 4920, 4928, 494, 4940, 4941, and 496 in the primary diagnosis; (2) ICD-9-CM codes: 4660 and 490 must combined with a secondary diagnosis code of COPD) were identified [12, 13] . ED visits and hospital admissions were separately defined as an independent event that occurred during the outcome period for each patient. The follow-up duration was defined as the number of days from the date of the end of the COC period to the date of the first ED visit or hospitalization for COPD or asthma, respectively. If a patient had no ED visits or hospital admissions in this period, then that the patient was censored at the end of the outcome period. We applied the Bice and Boxerman COC index (COCI) [32] as our primary dependent variable. COCI represents the concentration of visits to individual physicians and has been widely employed in studies using claim data sets [16] [17] [18] [19] [20] . In addition, the index is suitable for a higher number of outpatient visits [33] because it is less sensitive to the number of physician visits [34] . The COCI score, ranging from 0 to 1, measures the dispersion of contact between patient and physician, and the value close to 1 that represents greater COC. The general formula is Where N is the total number of physician visits, n i is the number of visits to the physician, and k is the total number of physicians. The total number of physician visits (N) and the number of visits to a given physician (n i ) included ambulatory claims (clinic or hospital physician visits) due to asthma or COPD as the principal diagnosis. COCI was categorized into three level (low (COCI < 0.3), medium (0.3 ≤ COCI<1) and high (COCI = 1)) [16, 35] . A number of control variables were considered in our study. Patients' characteristics at baseline were age, gender and insurance premiums (low: 12 outpatient visits for COPD or asthma in a calendar year should be more diligent regarding follow-up care. Those patients had a significant risk increase for ED visits and hospitalizations of COPD or asthma than those having 3-12 outpatient visits. The increased tendency in the rates of ED visit and admission for COPD or asthma in those patients with decreased COC levels further supports that COC might play a much more crucial role for those patients to reduce the risk of ED visits and admissions. Many indices were developed to evaluate COC such as COCI, Herfindhal Index, usual provider of care index (UPC), and Sequential Continuity of Care Index (SECON) in claims data [47] . COCI reflects the dispersion of contact between patients and physicians [32] . This index identifies visit concentration of a patient with each physician. The UPC, a density measure, focuses on the number of visits to the most frequently visited physicians, which cannot recognize whether patients reduce their visits or change healthcare providers frequently [48] . SECON determines the sequences of change in healthcare process, but it is limited to the detection of non-sequential issues [49] . The COCI, a most useful and common index, is adopted to measure for COC [50] . Additionally, compared with UPC and SECON, the COCI is suitable for a higher number of outpatient visits [33] because it is less sensitive to the number of physician visits [34] . Therefore, COCI was selected as our primary dependent variable to measure COC. In light of the effect of the P4P program of asthma care and COC is vital for COPD patients from our previous studies [20, 23] , the Taiwan NHI Administration has launched the COPD P4P program in 2017. ACO, which does not describe a single disease entity, likely includes patients with several different forms of airways phenotypes caused by a range of different mechanisms. However, there is limited evidence to investigate in regard of treating ACO [2] . This study used comprehensive data from a health care system with free access policy to provide the evidence that better COC is related to lower risk of ED visits and hospitalization for COPD or asthma among elderly ACO patients. Therefore, we suggest that policymakers could consider designing a program that focused on ACO to improve COC and disease-controlling ability. This study has some limitations. First, the claims data did not include clinical data such as peak expiratory flow or forced expiratory volume to identify ACO [40] . Therefore, we used ICD-9-CM codes to define ACO as done in previous studies [4, 6, 29] . Second, the claims data lacks personal anthropometric measurements such as body mass index [39, 40] and smoking status [39, 51, 52] that may affect healthcare outcomes. Therefore, we adopted ED visits and admissions for COPD or asthma as well as the frequency of ambulatory visits for COPD or asthma in the COC period as proxies for disease severity [9, 52] ; also, we used patients' history of hypertension and diabetes and CCI as proxy indicators to present each patient's health status [5, 40] . Third, we used insurance premiums as an indicator to represent patients' socioeconomic status because the database could not provide patients' educational level or household income, which may affect the measurement of COC and outcomes [8] . Forth, we excluded patients who had less than three outpatient visits during the COC period, which might limit the generalizability for elderly patients with ACO. Finally, claim-based COC measurements could offer only limited aspects of information regarding COC [33, 53] and fails to incorporate the view of COC from patient-centered point of view [54, 55] because measures are based on the patterns of patient encounters during a time period. Although some limitations exist, the study has several strengths. First, this retrospective cohort study presents stronger evidence of the association between COC and healthcare outcomes than a cross-sectional study. Second, the dataset are highly representative of the entire population because 99% of Taiwanese residents enroll in the NHI program which may reduce selection bias. Third, using the nationwide electronic health records (EHR) databases can reduce the effect of recall bias and make our results more informative. Fourth, this study focuses on ACO instead of either COPD or asthma, and provides empirical evidence of an adverse association between COC and the risk of ED visit and hospital admission. Fifth, this study adopted COCI which is the most commonly used index to measure the degree of coordination required between different providers during a clinical episode. Finally, the COCI was measured by continuity relationship between patients and physician, which may provide information that is superior to that obtained from measurements between patients and institutions. An adverse association exists between COC and the risk of ED visits and hospital admissions for COPD or asthma among elderly patients with ACO in Taiwan. Elderly ACO patients with higher COC were more likely to have lower risk of ED visits and hospital admissions for asthma or COPD. Therefore, elderly ACO patients should be encouraged to concentrate on few physicians they trust and to develop a long-term relationship with these physicians to enhance disease management. Policymakers could initiate P4P programs of ACO to motivate physicians and patients to enhance COC and disease control for reducing the ED visit and hospitalization. None declared. The asthma-COPD overlap syndrome Global Initiative for Asthma and Global Initiative for Chronic Obstructive Pulmonary Disease. Diagnosis and initial treatment of asthma, COPD, and asthma-COPD overlap Epidemiology of asthma-chronic obstructive pulmonary disease overlap (ACO) Burden of asthma and COPD overlap (ACO) in Taiwan: a nationwide population-based study Burden of concomitant asthma and COPD in a Medicaid population Disease burden of patients with asthma/COPD overlap in a US claims database: impact of ICD-9 codingbased definitions Cost and utilization of COPD and asthma among insured adults in the US The coexistence of asthma and chronic obstructive pulmonary disease (COPD): prevalence and risk factors in young, middle-aged and elderly people from the general population Medical utilization and cost in patients with overlap syndrome of chronic obstructive pulmonary disease and asthma The asthma-COPD overlap syndrome: a common clinical problem in the elderly Global strategy for the diagnosis, management and prevention of chronic obstructive lung disease 2017 report ACO #9 -prevention quality Indicator (PQI): ambulatory sensitive conditions admissions for chronic obsructive pulmonary disease (COPD) or asthma in older adults Continuity of care: a multidisciplinary review The association between continuity of care in the community and health outcomes: a population-based study Continuity of care for elderly patients with diabetes mellitus, hypertension, asthma, and chronic obstructive pulmonary disease in Korea Effect of continuity of care on emergency department visits in elderly patients with asthma in Taiwan Continuity and the costs of care for chronic disease Clinic continuity of care, clinical outcomes and direct costs for COPD in Sweden: a population based cohort study Effects of long-term high continuity of care on avoidable hospitalizations of chronic obstructive pulmonary disease patients Continuity of care and its effect on readmissions for COPD patients: a comparative study of Norway and Germany STROBE-compliant article: is continuity of care associated with avoidable hospitalization among older asthmatic patients? Medicine Continuity of care and avoidable hospitalizations for chronic obstructive pulmonary disease (COPD) National Health Insurance Administration. National Health Insurance annual report 2016-2017. Taipei Taiwan Ministry of Health and Welfare Reflections on the 20th anniversary of Taiwan's single-payer National Health Insurance System Severe acute respiratory syndrome-Taiwan Identifying individuals with physcian diagnosed COPD in health administrative databases Asthma-chronic obstructive pulmonary disease overlap syndrome associated with risk of pulmonary embolism Understanding asthma-chronic obstructive pulmonary disease overlap syndrome The association between continuity of care and outcomes: a systematic and critical review A quantitative measure of continuity of care Measuring care continuity: a comparison of claims-based methods Measures of continuity of care: a register-based correlation study Does continuity of care matter in a health care system that lacks referral arrangements? The effect of continuity of care on emergency room use for diabetic patients varies by disease severity Increased continuity of care associated with decreased hospital care and emergency department visits for patients with asthma Characteristics and prevalence of asthma/chronic obstructive pulmonary disease overlap in the United States Characterisation of the overlap COPD-asthma phenotype. Focus on physical activity and health status Coding algorithms for defining comorbidities in ICD-9-CM and ICD-10 administrative data Partial residuals for the proportional hazards regression model A longitudinal examination of continuity of care and avoidable hospitalization: evidence from a universal coverage health care system Continuity of care and the risk of preventable hospitalization in older adults Effects of continuity of care on emergency department utilization in children with asthma Does pay-for-performance benefit patients with multiple chronic conditions? Evidence from a universal coverage health care system Interpersonal continuity of care and care outcomes: a critical review Continuity of care in a university-based practice Measuring provider continuity in ambulatory care: an assessment of alternative approaches How important is continuity of care? Asthma and COPD overlap syndrome (ACOS): a systematic review and meta analysis Comorbidome, pattern, and impact of asthma-COPD overlap syndrome in real life How should continuity of care in primary health care be assessed? Validation of a generic measure of continuity of care: when patients encounter several clinicians What do we know about patients' perceptions of continuity of care? A meta-synthesis of qualitative studies The authors thank to the National Health Research Institutes for providing the National Health Insurance Research Database.