Waltham, MA, USA: Brandeis University (doctoral dissertation).
Brandeis University, Waltham, MA, USA
Health care organizations must find creative ways to achieve cost-efficient, high-quality primary care for their elderly patients. This dissertation research looks at the effect of a senior health center (SHC) model within a vertically integrated healthcare network on the delivery of preventive and chronic illness care services to the elderly. In particular, this dissertation research has three aims: 1. To evaluate whether a periodic health examination (PHE) correlates with provider recommendations for evidence-based preventive services (mammogram and colonoscopy screening, pneumonia vaccination), and identify how providers are impacted by the lack of reimbursement by Medicare for the PHE. 2. To evaluate whether there is a significant difference between the SHC model (where care is managed by a team of interdisciplinary geriatric providers) and the traditional PC model (where care is managed by a single provider specializing in internal medicine or family medicine) in the receipt of referrals for recommended ancillary care services for chronic illness. 3. To evaluate whether a healthcare network can benefit from the SHC investment by redirecting patients’ use of outpatient and inpatient services into the network, replacing services that were previously referred outside of the network.
Data used are drawn from the healthcare network’s electronic medical records of patients aged 65 and above for the period of 2000-07 (n=10,319 patients), along with face-to-face interviews of nine primary care providers in the network. Aim 1 of the dissertation is analyzed by mixed methods, consisting of fixed-effects regression (n=6,466), logistic regression (n=10,319), and qualitative analysis of provider interviews (n=9). Aim 2 of the dissertation is evaluated using a Generalized Estimating Equation (GEE) method; the sample used consists of traditional PC patients randomly selected to match with SHC patients in the post SHC implementation periods by age, gender, and health status (n=2,050). Aim 3 of the dissertation is measured using the Tooze correlated two-part model with a difference-in-differences regression format; the sample used is a pretest-posttest sample with rolling enrollment date and random selection (n=966).
Aim 1 results demonstrate that: 1) Patients who obtained a PHE had 2.17 (p0.0001) times greater odds, 1.54 (p0.0001) times greater odds, and 1.1 p=0.0226) times greater odds of obtaining a provider recommendation for mammogram screening, colonoscopy screening, and pneumonia vaccination, respectively. 2) Patients who obtained a PHE visit in the first visit year had 1.21 (p<0.0001) times greater odds of obtaining a provider recommendation for colonoscopy screening. Result of the effect of a PHE visit in the first visit year on provider recommendation for pneumonia vaccination is nonsignificant.
Qualitative data suggest that SHC providers (whose salaries are fixed, and who adopt a culture offering longer average appointment times for seniors than traditional PC providers) more readily integrate Medicare-uncompensated PHE services in routine visits than do traditional PC providers (who are paid based on a productivity scale that depends on the volume of visits, visit type, and intensity level of each office visit). To accommodate the lack of PHE, some traditional PC providers at the network either: 1) schedule patients for a longer follow-up visit in order to have the time to include some of the outstanding PHE services, and charge for a more complex visit than they normally would have charged; or 2) schedule more frequent follow-up visits to allow opportunities to cover the unmet PHE services in addition to follow-up care. In both cases, Medicare is indirectly paying for PHE services, despite the policy of not covering for a designated annual PHE visit.
Aim 2 results show that SHC patients diagnosed with osteoarthritis/stroke and depression/dementia were 9.7 percentage points (p0.001) and 14.4 percentage points (p0.001) more likely than traditionalPC patients to receive a referral to PT/OT and to neuopsychological evaluation, respectively.
Aim 3 results reveal that, in general, the healthcare network benefited from the SHC investment by retaining referrals within the network instead of referring patients outside of the network. The healthcare network experienced an increase in use of its therapy and nutrition education services, and radiology and MRI services, but not in its laboratory services and inpatient care. The SHC impact occurred more in the likelihood of use of services (part I of the model) than in the intensity of use per user (part II of the model). On average, SHC patients were more likely than traditional PC patients to use within-network therapy and nutrition education services, by about 8.1 (p<0.05) to 17.6 percentage points (p0.01) higher, depending on the post-period assessed. SHC patients were also more likely than traditional PC patients, on average, to use within-network MRI and radiology services, by about 9.3 (p0.05) to 20.6 (p0.01) percentage points higher, depending on the post-period assessed. (Abstract shortened by UMI.)
Please log in/register to access.