Extended Office Hours and Health Care Expenditures: A National Study

INTRODUCTION

The Joint Principles of the Patient-Centered Medical Home emphasize the importance of offering patients access to a usual source of health care during extended (eg, evening and weekend) hours.¹ Among several desired outcomes of offering extended access, one goal in the context of the health care financing crisis is to reduce unnecessary health care expenditures.² Offering patients extended access to their usual source of care might afford earlier diagnosis and treatment of serious health conditions, potentially preventing catastrophic events, and reducing expenditures. Extended office hours might also limit unnecessary patient exposure to higher-acuity (eg, emergency) care and, therefore, to discretionary and potentially harmful testing, treatments, and hospitalizations^3–5 initiated by clinicians unfamiliar with their care, which might further reduce expenditures.

Few studies, however, have examined associations between extended hours and health care use and expenditures. Most studies in this realm concerned emergency department use, with findings suggesting that extended hours may decrease emergency visits, but without consideration of effects on total health care expenditures.^6–10 An additional study found little impact of improved access to care on inpatient admissions and medical costs among chronically ill patients, but again did not examine effects on total expenditures.¹¹ The generalizability of these studies is uncertain, as they involved small local samples of patients, some with selected health conditions.

In addition, no studies have explored associations between extended office hours and mortality. This is an important research gap, as potential cost savings alone should not drive practice redesign, and competing influences of extended access on mortality may exist. Although extended access to a familiar clinician may foster more appropriate care, thereby reducing mortality, extended access could divert some patients with serious acute illnesses from emergency and hospital settings, delaying necessary emergency care and increasing mortality.

To address these issues, we analyzed longitudinal data from the Medical Expenditure Panel Survey (MEPS) for the years 2000–2008,¹² and, for respondents enrolled in 2000–2005, linked data to the National Death Index (NDI) though 2006.¹³ We examined associations between reported consistent access or lack of access to extended office hours via a usual source of care in 2 successive years (year 1 and year 2) and year 2 total health care expenditures and subsequent mortality. We hypothesized that self-reported access to extended office hours would be associated with lower health care expenditures, without an adverse effect on mortality. We also explored mechanisms of the association between extended access and expenditures in analyses progressively adjusting for subcategories of year 2 health care use and expenditures: (1) emergency department expenditures and inpatient expenditures, (2) number of office visits and number of prescription medications, (3) prescription medication expenditures, and (4) office visit–related expenditures.

METHODS

The MEPS is an annual national survey of health care use and costs in the US civilian, noninstitutionalized population, using an overlapping panel design.¹² Individual data are collected over a 2-year period through 6 interviews. The MEPS Household Component includes information on respondent health care features, sociodemographics, health insurance, and health care expenditures. A self-administered questionnaire in both years includes items on chronic health conditions and health status. The full-year response rate varied from 65.8% to 59.3% for the 9 panels of data (2000–2008) we used.¹² Our analytic sample included all persons aged 18 years and older on entry reporting a usual source of care in both years, and either access to extended office hours in both years or no such access in both years.

The MEPS Household Component sample is a sub-sample of households included in the previous year’s National Health Interview Survey (NHIS), conducted annually by the National Center for Health Statistics. The NHIS is linked to death certificate data in the NDI, a central computerized index of US death record information on file in state vital statistics offices, in turn permitting linkage to the MEPS.¹³

RESULTS

A total of 54,624 eligible adults aged 18 to 90 years entering the MEPS cohorts between 2000 and 2007 reported a usual source of care in both years; 43,484 (79.6%) had complete baseline data. Of these, 33,269 (77.5%, population weighted) reported the same access to extended hours in both years (κ for between-year agreement, 0.77; agreement expected by chance, 0.53) and composed the analytic sample.

Table 1 summarizes the characteristics of the analytic sample by whether or not the respondent reported access to extended office hours in both years. Compared with respondents reporting no such access, respondents reporting access were younger, had higher incomes and more education, were more likely to reside in urban areas and the Northeast, were more likely to have private insurance, and had better health status and fewer health conditions.

Table 2 shows unadjusted health care use and expenditures by reported access to extended hours. Those reporting access to extended hours had less use and lower related expenditures for all subcategories of use (office visits, prescription medications, emergency department visits, and hospitalizations).

Table 3 shows the adjusted associations between respondent characteristics and the logarithm of year 2 expenditures (model R² = 42%); 6.7% (population weighted) had no expenditures. As compared with reported lack of access to extended hours, access to extended hours was associated with significantly lower year 2 expenditures (adjusted parameter estimate, −0.11, 95% confidence interval [CI] −0.14 to −0.07; P <.01). As shown in Table 4, this association translated to 10.4% lower year 2 expenditures (95% CI, 7.2%–13.4%; P <.01) in the group reporting access to extended hours. Supplementary adjusted analyses exploring possible pathways for the relationship between extended office hours and expenditures revealed that reported access to extended hours was associated with a significantly lower level of year 2 emergency department use (1.9% fewer visits; 95% CI, 0.8%–3.7%; P = .04), but not with year 2 hospitalizations.

Table 4 also shows the change in the magnitude of the association between reported extended hours access and total health care expenditures when additionally adjusted for year 2 health care use. The association was not substantively attenuated by added adjustment for year 2 emergency department and inpatient expenditures or for total number of office visits and number of prescription medications. The association was attenuated, however, with added adjustment for prescription expenditures and attenuated further with adjustment for office visit–related expenditures.

A total of 833 respondents (3.7%) had died by the end of 2006, 191 (2.3%) among those reporting access to extended hours in both years, and 642 (4.0%) among those reporting no extended access in both years. Mortality was not associated statistically with reporting extended hours among respondents enrolled in 2000–2005 panels (adjusted hazard ratio = 1.11; 95% CI, 0.92–1.35; P = .28; N = 22,766; 833 [3.7%] died). Associations between reporting extended hours and expenditures were consistent in all supplementary analyses, including analyses using generalized linear models with Poisson distributions, and adjusting for site of usual source of care and health risks (body mass index and smoking).

DISCUSSION

Patients reporting in both study years that their usual source of care offered evening and weekend hours had lower year 2 total health care expenditures than those consistently reporting no evening and weekend access, without apparent adverse effects on mortality. The findings stemmed from analyses adjusted for numerous potential confounders, including sociodemographics and health insurance status plus an array of variables capturing comorbidity (year 2 mental and physical health status, self-rated health, and chronic conditions, and year 1 health care expenditures and use). Indeed, the model explained 42% of the variance in total expenditures, more than most expenditure prediction models, typically including less comorbidity adjustment.¹⁸

Furthermore, in analyses progressively adjusting for aspects of year 2 health care use, only adjustment for year 2 prescription drug expenditures and office visit–related expenditures (but not number of prescriptions or office visits) substantively attenuated the relationship between extended hours and total expenditures. Although our study design does not permit formal testing of causal pathways, 2 interpretations of this finding seem plausible. First, clinicians in practices offering extended access may tend to provide more cost-conscious care in general, including a proclivity for prescribing less expensive (eg, generic) medications and less discretionary test ordering during office visits. A patient-centered approach has been previously demonstrated to be associated with decreased use of and charges for medical services.¹⁹ Other work suggests significant independent effects of clinicians on generic drug and discretionary care use.^20–22 Extended access may thus represent a marker of primary care practices with a cost-conscious approach. This finding may be useful to health administrators and policy makers interested in relatively simple ways of identifying such practices—a seemingly important aim given the health care financing crisis.² Additionally, or alternatively, practices offering extended access may attract patients less likely, from a dispositional standpoint, to request brand name medications and discretionary testing. Studies designed to examine these and other mechanisms would be helpful.

Of note, although patients reporting access to extended hours were less likely to visit the emergency department in year 2, reduced emergency use did not account for the lower total health care expenditures in year 2 associated with extended access, as we and others^6–10 had contemplated. With hindsight, this is not a particularly surprising finding, given that emergency department–related expenditures represented only a small proportion of total year 2 health care expenditures (Table 2). The association between extended hours and total expenditures also did not appear attributable to another factor we considered salient—clinician familiarity with the patient (eg, continuity)—given the lack of meaningful attenuation when adjusting for year 2 office visits. Rather, as noted previously, adjustment for year 2 office visit expenditures attenuated the relationship, suggesting clinician practice style (eg, proclivity for discretionary testing) may influence total expenditures more than patient continuity. Again, however, further studies are needed to test these hypotheses.

Although health care costs are an important consideration in the current climate, patient outcomes remain the most important criteria for assessing quality of care, as some cost-saving care approaches may foster unacceptably worse health outcomes.²³ Specifically, the availability of extended office hours could increase mortality by diverting patients in need of higher-acuity care away from emergency departments and hospitals. In this context, although reported access to extended hours was not associated with decreased mortality, that it was not associated with increased mortality is reassuring.

Limitations of this study include the MEPS reliance on patient survey data. Self-reports of having a usual source of care may overstate the respondent’s actual connections to a primary care clinician. There may also be some bias in recall, in that patients recalling the availability of extended hours may also be those likely to generate lower expenditures. Our approach of including only respondents reporting the same access to extended hours in both survey years probably reduced this problem; however, it is likely that some misclassification bias remained, and this bias may have resulted in underestimation of the association between extended hours and total expenditures. In addition, we used self-reported health care use and standardized expenditures in our analyses, rather than more direct measures of visits, tests, or claims. We were also unable to adjust for some variables that might plausibly influence health care expenditures beyond those included in our models, such as specific type of health care system (eg, vertically integrated health maintenance organization vs stand-alone primary care office) and health insurance plan, and health care quality (eg, National Committee for Quality Assurance indicators).

In conclusion, respondents consistently reporting access to a regular source of care offering evening and weekend hours in both study years had significantly lower year 2 total health care expenditures, but not mortality. The association was substantially attenuated by adjusting for year 2 prescription medication and office visit–related expenditures, but not number of medications and office visits or emergency department and inpatient use. Although studies designed to formally test this hypothesis are needed, extended hours may be associated with relatively judicious use of primary health care resources (eg, more generic medication prescribing, less discretionary testing), with no apparent adverse effects on survival.