Bergmark RW, Ishman SL, Scangas GA, Cunningham MJ, Sedaghat AR. Insurance Status and Quality of Outpatient Care for Uncomplicated Acute Rhinosinusitis. JAMA Otolaryngol Head Neck Surg. 2015;141(6):505-511. doi:10.1001/jamaoto.2015.0530
Previous work suggests an association between insurance status and location of presentation (emergency department vs outpatient clinic) for evaluation of uncomplicated acute rhinosinusitis (ARS).
To investigate whether the quality of outpatient care for ARS likewise differs based on insurance status.
Design, Setting, and Participants
Cross-sectional study of 13 680 145 pediatric and adult patients from the 2009 and 2010 National Ambulatory Medical Care Survey diagnosed with uncomplicated ARS at an outpatient care facility.
Health insurance status.
Main Outcomes and Measures
The primary outcome measures were continuity of care with the patients’ primary care physician (PCP) and time spent with a physician, which were chosen as proxies for quality of care. We evaluated associations between insurance status and these quality measures while controlling for clinical, demographic and socioeconomic patient characteristics, and outpatient practice setting.
Most patients (76.4%) had private insurance vs Medicare (12.3%), Medicaid (8.6%), or self-pay (2.8%). There was no association between insurance status and presentation of patients to their PCP. Physicians spent more time with Medicaid patients compared with patients with private insurance (β = 4.59; P = .01), independent of other factors. Provision of health education (β = 4.42; P < .001), necessity of a follow-up visit (β = 3.20; P = .002), and increasing patient age (β = 0.07; P = .01) were associated with longer visits. In multivariate analysis, living in a medium or small metropolitan area was associated with higher likelihood of being seen by one’s own PCP than living in a large metropolitan area (odds ratio, 6.37; 95% CI, 2.13-19.05; P = .001).
Conclusions and Relevance
This study did not identify any quality of care issues with respect to insurance coverage and primary care encounters for patients with ARS. In contrast to expectations, patients with Medicaid had longer outpatient physician visits and were equally likely to see their own PCP compared with patients with private insurance or Medicare.
Health care disparities have been well described in the medical literature. Insurance status has been associated with disparate health care utilization and outcomes. Numerous studies have shown strong associations between Medicaid insurance or no insurance and difficulty obtaining services as well as adverse health outcomes.1- 5 Provision of adequate health insurance and increasing access to care are paramount in the elimination of socioeconomic disparities in health care.6,7 One area of disparate health care utilization with major economic implications is the differential utilization of primary care vs emergency department services for nonemergent conditions. Previous studies have found Medicaid or lack of insurance to be associated with limitations to primary care access and increased use of emergency department care for nonemergent conditions, which could be appropriately managed in the office setting.8- 10 Such is true for patients with acute rhinosinusitis (ARS) with Medicaid and self-pay insurance statuses who have been shown to preferentially use emergency department care over primary care compared with patients who have private insurance or Medicare.11
Acute rhinosinusitis, a common sequela of upper respiratory tract infections,12 is one of the most commonly diagnosed conditions, with an estimated 20 million cases occurring annually in the United States.13 Due in large part to this high incidence, ARS has a considerable economic impact on health care expenditures. In the United States, the direct costs of sinusitis management are estimated at $3 billion per year inclusive of initial medications and antibiotic treatment failures, ancillary tests and procedures, and outpatient and emergency department visits.12,13
Given previous work suggesting that disparities in quality of outpatient care may exist based on insurance status,1,3 we hypothesized that disparities in the quality of outpatient care might be related to preferential emergency department use for uncomplicated ARS by patients with Medicaid or self-pay insurance status. Herein we use continuity of care with an outpatient primary care physician (PCP) and time spent by the physician with the patient as metrics for quality of care. We examine whether differences exist in these metrics for uncomplicated ARS visits in outpatient clinics based on insurance status.
Ethical review and written informed consent were waived by the Massachusetts Eye and Ear Infirmary institutional review board for this retrospective database analysis.
All adult and pediatric outpatient primary care visits from the 2009 and 2010 National Ambulatory Medical Care Survey (NAMCS) were identified by searching for patients visiting an internist, general or family practitioner, or pediatrician included in the databases. Visits for ARS were defined as those with at least 1 International Classification of Diseases, Ninth Revision (ICD-9) code for acute sinusitis (461.0, 461.1, 461.2, 461.3, 461.8, and 461.9) assigned to any of the 3 diagnoses included in the databases. The NAMCS is a national probability sample survey of patient visits to non–federally employed, office-based physicians practicing direct patient care in the United States that is maintained by the Centers for Disease Control and Prevention. Excluded from analysis were patients with ICD-9 codes for ARS complications including preseptal or periorbital cellulitis (373.13), orbital cellulitis or abscess (376.0, 376.01, 376.02, and 376.03), intracranial complications of intracranial abscess (324.0), and cavernous sinus thrombosis (325.0).
The primary outcome measures for this study of ARS visits were (1) whether the visit was with the patient’s PCP, reflecting continuity of care, and (2) the length of time spent by the physician with the patient. Time spent with the physician is a well-established component of patient satisfaction and quality of care. The Visit Specific Patient Satisfaction Questionnaire (VSQ) includes time spent with the physician as 1 of its 4 major outcome measures.14,15 The Press Ganey outpatient medical practice survey also includes time with the physician as a patient satisfaction and quality metric.16 Characteristics of the patient visit recorded included whether the patient was seen by a physician, whether the patient had been seen at this practice before, whether health education was given, and the disposition of the patient (return visit scheduled, referred to another physician, or referred to an emergency department). Documented characteristics of the clinical practice included whether it was a solo practice, whether patients were seen on weekends or nights, and the type of office setting (private practice, urgent care, community health center, non–federal government clinic or health maintenance organization).
Patient characteristics of age, sex, and race were extracted in addition to the primary payer (ie, insurance type) listed for the visit. A temperature greater than 37.8°C (hereinafter, 100.0°F) was also recorded and used as an indicator of ARS severity. A reflection of patient complexity was taken to be the number of associated diagnosed chronic conditions defined by the NAMCS as arthritis, asthma, cancer, cerebrovascular disease, chronic renal failure, congestive heart failure, chronic obstructive pulmonary disease, depression, diabetes mellitus, hyperlipidemia, hypertension, ischemic heart disease, obesity, and osteoporosis. Characteristics of the patients’ home zip code were extracted including (1) median income level for the zip code, (2) the percentage of adults with a bachelor’s degree in that zip code, and (3) metropolitan status of that zip code (as defined by criteria from the US Office of Management and Budget).
All analysis was performed with the statistical software package R (http://www.r-project.org). Weighted associations between predictor variables and time spent with the physician (by linear regression) and whether the patient was seen by his or her primary care physician (by logistic regression) were performed with the Regression Modeling Strategies (rms) package.17 Missing data were omitted. Quality control was performed by confirming the minimum numbers of data entries for each category as well as calculation of relative standard errors (SEs) to ensure reliability of results, as recommended by the Centers for Disease Control and Prevention. Univariate logistic regression was performed for each predictor variable. Multivariate analysis was performed using all predictor variables with P ≤ .100 in the univariate regression. In the multivariate model, significant predictors were identified via backward elimination, using a P value cutoff of 0.100.17 Cross-validation was performed through bootstrapping of the data set using the validate() function from the rms package over 100 iterations. For each variable retained in the final model, a P value and a log–odds ratio were calculated; P < .05 was considered significant.
A total of 13 680 145 outpatient clinic visits for ARS over 2009 and 2010 were identified in the NAMCS. The clinical and demographic characteristics of these patients were tabulated (Table 1). Most patients had private insurance (76.4%) vs Medicare (12.3%), Medicaid (8.6%), or self-pay insurance status (2.7%). The mean (SD) patient age was 40.1 (20.4) years, and 57.3% of the patients were female. Only 1.8% of patients had a temperature over 100.0°F at the time of the ARS clinic visit. Almost all patients had either no chronic medical conditions (49.9%) or 5 or fewer chronic conditions (49.7%). Most patients identified their race as white (76.5%) vs black (10.1%), Hispanic (9.8%), or other (3.6%). Patients living in zip codes in the poorest income quartile were underrepresented, accounting for 15% of patients. Half of all patients (50.9%) lived in a large metropolitan area, 38.0% in a medium or small metropolitan area, and 11.1% in a rural area.
Many of the patient demographic characteristics for ARS visits were similar to those of the overall greater than 1 billion primary care visits captured in the 2009 and 2010 NAMCS (Table 1). Notably, however, insurance status for ARS visits was skewed toward private payers, whereas insurance status for primary care visits in general consisted of a greater proportion of public insurances (Medicare and Medicaid). Lower percentages of uninsured and Medicaid patients in the patient population presenting with ARS are also consistent with lower percentages of patients in the lowest income levels presenting with ARS compared with all primary care visits.
The characteristics of the practices managing these patients with ARS were determined (Table 2). The vast majority of patients were evaluated (1) by a physician as opposed to a nurse or physician assistant during their visit (99.1%), (2) at a clinical practice where they had previously been seen (93.5%), and (3) by their own PCP (84.0%). Most patients (84.3%) saw a physician in a private practice setting. A quarter of physicians were in a solo practice. A high proportion of physicians (42.7%) offered to see patients on weekends or nights. Physicians spent a mean (SD) of 17.1 (8.1) minutes with each patient with ARS. In 31.6% of ARS cases, physicians reported giving health education to the patient, and 35.8% of patients were scheduled for a return appointment.
Patients’ characteristics were tested for association with being seen by their own PCPs and reported as odds ratios (ORs) and 95% CIs (Table 3). The number of patient entries with “self-pay” insurance did not meet the minimum quality control standards for reliability set by the Centers for Disease Control and Prevention; these patients were therefore excluded from further analysis. On univariate analysis, increasing age was associated with seeing one’s own PCP (OR, 1.02; P = .002). Categorization of age into pediatric vs adult did not result in a better-fitting regression. Other patient characteristics found to be associated with seeing one’s own PCP were the number of chronic conditions (OR, 1.90; P = .01) and residence in a medium or small metropolitan area (OR, 7.60; P < .001) or nonmetropolitan area (OR, 10.71; P = .03). Compared with patients with private insurance, there was no less likelihood of seeing one’s own PCP among patients with Medicare (OR, 1.50; P = .46) or Medicaid (OR, 3.62; P = .16) insurance status. Multivariate analysis confirmed a statistically significant association between seeing one’s own PCP and residence in a medium or small metropolitan area (OR, 6.37; P = .001).
We also utilized the amount of time spent with a physician for an ARS visit as a metric for the quality of care, testing for associations with characteristics of the patient and the clinical practice (Table 4). Univariate regression found greater time spent with the physician for ARS was associated with Medicaid insurance (β [coefficient of regression] = 5.42; P = .003) compared with private insurance. Patient age (β = 0.05; P = .023) was associated with greater time spent with the physician, and no better-fitting regression was found by categorization of age into pediatric vs adult. Also associating with greater time spent with the physician on univariate analysis were the total number of chronic conditions (β = 1.40; P = .001), nonwhite race (β = 2.32; P = .04) compared with whites, the provision of health education by the physician (β = 5.19; P < .001), solo practice (β = 3.95; P < .001) and when a follow-up appointment was scheduled (β = 4.59; P < .001). Multivariate regression confirmed that greater time spent with the physician for ARS was associated with Medicaid insurance (β = 4.59; P = .01) as well as older age (β = 0.07; P = .01), provision of health education (β = 4.42; P < .001), and when a follow-up appointment was scheduled (β = 3.20; P = .002).
Previous work has suggested that patients’ disparate access to health care and quality of outpatient care may be based, in part, on their health insurance status.8,9,18 Such disparities in primary care may promote unnecessary emergency department utilization, which represents a potentially significant source of excessive health care cost.10,19,20 We previously found both Medicaid and self-pay insurance statuses to be associated with emergency department visits for uncomplicated ARS,11 a nonemergent condition more appropriately managed in an outpatient primary care clinic setting. In this study, we hypothesized and investigated whether disparate quality of primary care for patients with Medicaid or self-pay insurance status may provide insight into the disparate emergency department utilization for uncomplicated ARS by these patients. We did not, however, identify such disparities based on our quality of care metrics. Unexpectedly, compared with patients with Medicare or private insurance, patients with Medicaid or self-pay insurance status had longer visits with their physician and were equally likely to see a physician and for that physician to be their own PCP. They were also equally likely to be seen in a practice where they had been cared for previously, suggesting nondisparate continuity of care. These results are independent of the patients’ comorbidities and income level.
Increasing interest in studying access to and quality of outpatient care visits for conditions such as ARS has stemmed in part from concern over excessive emergency department utilization. Indeed, prior work has shown that patients with Medicaid or no insurance (self-pay) are more likely to present to the emergency department for acute care of outpatient- and primary care–treatable conditions.10,11 Potential reasons for this include lack of access to outpatient care, inability to find a physician or make an appointment, lack of trust in the outpatient setting, or the presence of pain and other clinical characteristics that the patient believed mandated emergency department presentation.21
Disparities in access and quality of health care based on insurance status have been well investigated in other medical fields. Across the United States, 31% of physicians do not accept Medicaid patients into their practice, compared with 18% for privately insured patients; higher state Medicaid-to-Medicare fee ratios have been found to correlate with higher Medicaid acceptance rates.22 It is not surprising then that Medicaid and lack of insurance have been associated with lack of access to a PCP, even for urgent conditions that are more appropriately treated in the clinic setting.8,23,24 In a large study,8 only 34% of callers posing as Medicaid patients were able to make an urgent follow-up appointment vs 64% of callers posing as patients with private insurance. This disparity was even worse in an earlier study,25 which found that only 26% of private practices would agree to see a Medicaid patient within 2 working days for an urgent condition vs 60% for a privately insured patient. Although a third, more recent study26 found higher overall appointment rates, Medicaid patients continue to have diminished access and lower success in securing office appointments (58%) than private patients (85%).
If these previously reported difficulties with obtaining office appointments are a reflection of latent opinions toward insurance types, it is likewise possible that there may be differences in the quality of the outpatient clinic care experience for Medicaid patients vs patients with private insurance. For other diseases, it has been shown that lower socioeconomic status is associated with lower likelihood of receiving evidence-based care such as mammograms, Papanicolaou tests (aka, “pap smears”), and appropriate cardiovascular drugs.27 One review of socioeconomic and racial disparities in health care27 cited multiple associated factors including “health care affordability, geographic access, transportation, education, knowledge, literacy, health beliefs, racial concordance between physician and patient, patient attitudes and preferences, [and] competing demands including work and child care.”27(p2580) Patients with Medicaid, minority patients, and non–English speaking patients score more poorly in standard primary care outcomes measures.28 Moreover, qualitative studies have found that patients of lower socioeconomic status perceive their outpatient clinic options as more challenging to access, more expensive, and of lower quality than the emergency department.21
In this study, we used 2 measures of health care quality to determine if disparities existed for patients presenting to outpatient clinics for uncomplicated ARS: (1) whether the patient saw his or her own PCP as a proxy of continuity of care, and (2) how much time the physician spent with the patient. From these 2 metrics, we found that Medicaid or self-pay patients did not experience any obvious health care quality deficits.
The study had some limitations. Our results are based on data contained within the NAMCS. This data set, by definition, excludes patients without access to primary care. Likewise, patients who do not believe they are able to get adequate care in the outpatient clinic setting may not present to their PCP when they have an acute problem such as uncomplicated ARS and would therefore not be included in this data set. The NAMCS also does not include other potentially valuable quality measures such as language preference, use of interpreters, wait time, ease of making urgent appointments, copayment information, and other barriers that might limit access to outpatient care. As such, these metrics for quality of care for ARS could not be measured.
Regardless of insurance status, patients presenting to outpatient clinics for uncomplicated ARS almost always see a physician (generally their own PCP or someone in their practice) with no substantial disparities in the amount of time spent with the physician. These results do not suggest a reason for disparate emergency department use for uncomplicated ARS of patients with Medicaid and self-pay insurance status. While our results do not exclude the possibility that insurance-based disparities in outpatient clinic visits for ARS exist, any such disparities are likely too complex and subtle for the health care quality metrics selected for this investigation to detect.
Submitted for Publication: December 24, 2014; final revision received February 12, 2015; accepted March 9, 2015.
Corresponding Author: Ahmad R. Sedaghat, MD, PhD, Department of Otolaryngology–Head and Neck Surgery, Massachusetts Eye and Ear Infirmary, Harvard Medical School, 243 Charles St, Boston, MA 02114 (firstname.lastname@example.org).
Published Online: April 16, 2015. doi:10.1001/jamaoto.2015.0530.
Author Contributions: Dr Sedaghat had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Bergmark, Sedaghat.
Acquisition, analysis, or interpretation of data: Bergmark, Ishman, Scangas, Cunningham, Sedaghat.
Drafting of the manuscript: Bergmark, Ishman, Scangas, Sedaghat.
Critical revision of the manuscript for important intellectual content: Bergmark, Ishman, Scangas, Cunningham, Sedaghat.
Statistical analysis: Bergmark, Scangas, Sedaghat.
Study supervision: Ishman, Cunningham, Sedaghat.
Conflict of Interest Disclosures: None reported.
Previous Presentations: This work was presented as a poster at the American Rhinology Society Meeting at the Academy of Otolaryngology–Head and Neck Surgery Annual Meeting; September 20, 2014; Orlando, Florida.