William B. Weeks

THE ASSOCIATION BETWEEN USE OF CHIROPRACTIC CARE AND COSTS OF CARE AMONG OLDER MEDICARE PATIENTS WITH CHRONIC LOW BACK PAIN AND MULTIPLE COMORBIDITIES

OBJECTIVE The purpose of this study was to determine whether use of chiropractic manipulative treatment (CMT) was associated with lower healthcare costs among multiply-comorbid Medicare beneficiaries with an episode of chronic low back pain (cLBP). METHODS We conducted an observational, retrospective study of 2006 to 2012 Medicare fee-for-service reimbursements for 72326 multiply-comorbid patients aged 66 and older with cLBP episodes and 1 of 4 treatment exposures: chiropractic manipulative treatment (CMT) alone, CMT followed or preceded by conventional medical care, or conventional medical care alone. We used propensity score weighting to address selection bias. RESULTS After propensity score weighting, total and per-episode day Part A, Part B, and Part D Medicare reimbursements during the cLBP treatment episode were lowest for patients who used CMT alone; these patients had higher rates of healthcare use for low back pain but lower rates of back surgery in the year following the treatment episode. Expenditures were greatest for patients receiving medical care alone; order was irrelevant when both CMT and medical treatment were provided. Patients who used only CMT had the lowest annual growth rates in almost all Medicare expenditure categories. While patients who used only CMT had the lowest Part A and Part B expenditures per episode day, we found no indication of lower psychiatric or pain medication expenditures associated with CMT. CONCLUSIONS This study found that older multiply-comorbid patients who used only CMT during their cLBP episodes had lower overall costs of care, shorter episodes, and lower cost of care per episode day than patients in the other treatment groups. Further, costs of care for the episode and per episode day were lower for patients who used a combination of CMT and conventional medical care than for patients who did not use any CMT. These findings support initial CMT use in the treatment of, and possibly broader chiropractic management of, older multiply-comorbid cLBP patients.

Comparing Propensity Score Methods for Creating Comparable Cohorts of Chiropractic Users and Nonusers in Older, Multiply Comorbid Medicare Patients With Chronic Low Back Pain

OBJECTIVE Patients who use complementary and integrative health services like chiropractic manipulative treatment (CMT) often have different characteristics than do patients who do not, and these differences can confound attempts to compare outcomes across treatment groups, particularly in observational studies when selection bias may occur. The purposes of this study were to provide an overview on how propensity scoring methods can be used to address selection bias by balancing treatment groups on key variables and to use Medicare data to compare different methods for doing so. METHODS We described 2 propensity score methods (matching and weighting). Then we used Medicare data from 2006 to 2012 on older, multiply comorbid patients who had a chronic low back pain episode to demonstrate the impact of applying methods on the balance of demographics of patients between 2 treatment groups (those who received only CMT and those who received no CMT during their episodes). RESULTS Before application of propensity score methods, patients who used only CMT had different characteristics from those who did not. Propensity score matching diminished observed differences across the treatment groups at the expense of reduced sample size. However, propensity score weighting achieved balance in patient characteristics between the groups and allowed us to keep the entire sample. CONCLUSIONS Although propensity score matching and weighting have similar effects in terms of balancing covariates, weighting has the advantage of maintaining sample size, preserving external validity, and generalizing more naturally to comparisons of 3 or more treatment groups. Researchers should carefully consider which propensity score method to use, as using different methods can generate different results.

The United States Chiropractic Workforce: An alternative or complement to primary care?

BackgroundIn the United States (US) a shortage of primary care physicians has become evident. Other health care providers such as chiropractors might help address some of the nation’s primary care needs simply by being located in areas of lesser primary care resources. Therefore, the purpose of this study was to examine the distribution of the chiropractic workforce across the country and compare it to that of primary care physicians.MethodsWe used nationally representative data to estimate the per 100,000 capita supply of chiropractors and primary care physicians according to the 306 predefined Hospital Referral Regions. Multiple variable Poisson regression was used to examine the influence of population characteristics on the supply of both practitioner-types.ResultsAccording to these data, there are 74,623 US chiropractors and the per capita supply of chiropractors varies more than 10-fold across the nation. Chiropractors practice in areas with greater supply of primary care physicians (Pearson’s correlation 0.17, p-value < 0.001) and appear to be more responsive to market conditions (i.e. more heavily influenced by population characteristics) in regards to practice location than primary care physicians.ConclusionThese findings suggest that chiropractors practice in areas of greater primary care physician supply. Therefore chiropractors may be functioning in more complementary roles to primary care as opposed to an alternative point of access.

Over-REACHing Conclusions

To the Editor: In their analysis of the healthcare cost savings associated with implementation of the Research for Enhancing Alzheimer’s Caregivers Health (REACH) dementia care program, Nichols and colleagues present findings from two studies: REACH II, a prospective trial of 197 Medicare beneficiaries randomly assigned to control or REACH care; and REACH in the Department of Veterans Affairs (REACH VA), a retrospective cohort study comparing 93 veterans who received REACH within the VA system with 398 retrospectively propensity-scorematched veterans who did not. In REACH II, the authors found that REACH care resulted in 24.7% statistically nonsignificantly higher Medicare costs; in REACH VA, they found statistically significantly (P = .05) 33.6% lower VA care costs with the intervention that contracted to statistically nonsignificantly 23.3% lower costs when considering VA and Medicare costs. The authors offered two explanations for these discrepancies: either REACH II was underpowered to demonstrate cost savings, or REACH VA drew from beneficiaries enrolled in a program that is more integrated than the “more fragmented feefor-service care received by REACH II participants” (although the authors did not have data on the number of REACH II participants enrolled in integrated care models). There may be another explanation, should readers interpret the retrospective cohort study in light of the prospective, randomized, controlled study, instead of vice versa. From that vantage point, Medicare costs of REACH VA participants were also increased substantially—by

Association Among Opioid Use, Treatment Preferences, and Perceptions of Physician Treatment Recommendations in Patients With Neck and Back Pain

1,110 (or ~47%) from a baseline of

Association Between a Measure of Community Economic Distress and Medicare Patients’ Health Care Utilization, Quality, Outcomes, and Costs

2,367 (the difference between the 33.6% lower VA care costs (

Proportion of Decedents’ Expenditures Among Recent Reductions in Medicare Expenditures

5,421) and the total cost savings of 23.3% (

Ineffective Policies to Address the Opioid Epidemic

4,311) for VA plus Medicare costs). In both studies, the Medicare cost effect appeared to be in the same direction and of similar magnitude relative to baseline use. Why did VA costs drop so much in the REACH intervention group? A concern is that prior year use between the control and intervention groups was so different; the intervention group had 32.7% higher baseline VA costs (

Public Perceptions of Doctors of Chiropractic: Results of a National Survey and Examination of Variation According to Respondents' Likelihood to Use Chiropractic, Experience With Chiropractic, and Chiropractic Supply in Local Health Care Markets

16,136 vs

A proposal to improve health-care value in spine care delivery: the primary spine practitioner

12,156). As the authors note, these cost data are highly skewed; because of this, the authors log-transformed cost data in the difference-in-difference analysis, although they used t-tests, which assume a normal distribution, in their baseline cost analysis. Using simulated cost data based on the parameters provided in the article to run a Monte Carlo simulation and repeating the analysis 1,000 times, I found that, had prior-year VA spending data been log-transformed, the control and intervention groups would have been statistically significantly different at baseline. Why does this matter? Although the authors might argue that difference-in-difference analyses account for baseline differences, those differences support the possibility that regression to the mean could explain their findings and cause a type I error; REACH implementation could have been incidental to the intervention group’s high baseline costs returning to normal. Would that the authors had incorporated prior-year VA andMedicare costs into their matching algorithm. In light of the statistically nonsignificant but substantial increase in costs associated with REACH implementation in REACH II (in which the two Medicare populations had similar prior-year total Medicare costs (