VAMCs Fall Short in Behavioral Healthcare, Patient Satisfaction

By Brenda L. Mooney

CHICAGO — Beginning last fall, VHA’s hospitals were included on the Hospital Compare website maintained by the Centers for Medicare and Medicaid (CMS), allowing direct comparison with other, nonfederal facilities.

While it might come as a surprise to some of the VA’s most fervent critics, a new analysis suggested that VAMCs actually did quite well, especially in providing safe and effective medical care, as opposed to behavioral healthcare. The healthcare system still wouldn’t win any popularity contests, however.

A research letter published recently in JAMA Internal Medicine, pointed out that VA hospitals had better outcomes than non-VA hospitals for six of nine patient safety indicators (PSIs), with no significant differences for three others. VA hospitals also had better outcomes for all of the mortality and readmissions metrics.1

Less positive news came with the patient experience measures, according to the authors from Feinberg School of Medicine at Northwestern University. Non-VA hospitals posted better scores, overall, than VAMCs for nursing and physician communication, responsiveness, quietness, pain management and on whether the patient would recommend the hospital to others. As for behavioral health measures, non-VA hospitals did better on four of nine measures, while VA hospitals did better on one of nine measures.

“While concerns remain about the validity of some of the measures used in current public reporting of health quality, the available data suggests that VA hospitals have a similar or more favorable quality compared with non-VA hospitals,” the authors wrote. “On the other hand, these results suggest that VA hospitals should focus on improving certain aspects of patient experience and behavioral health. Hospitals can use these data to identify opportunities for quality improvement.”

The researchers noted that quality comparisons between VA and non-VA hospitals previously were inconclusive and had methodological limitations. “Given long-standing concerns about care at VA hospitals,4 our objective was to compare available outcome, patient experience, and behavioral health measures between VA and non-VA hospitals,” they explained.

For the study, hospital-level data were obtained for 129 VA and 4,010 non-VA hospitals through the CMS Hospital Compare website. Non-VA hospitals were classified as major teaching other teaching, community, specialty and critical access hospitals.

The facilities were compared on 17 outcome measures—including mortality and readmissions—10 patient experience measures and nine behavioral health measures.

“In this evaluation of the most recent, comprehensive public reporting of VA hospital quality that can be compared with non-VA hospitals, we found that VA hospitals performed better than non-VA hospitals for most outcome measures, but VA hospitals performed worse on certain patient experience measures and behavioral health measures,” study authors wrote. “A recent systematic review was inconclusive about the quality of care provided at VA hospitals vs. non-VA hospitals. In that analysis, VA hospitals generally performed better than non-VA hospitals with respect to safety and effectiveness of care; however, most of the studies in this review examined only a subset of patients, had a smaller sample size of non-VA Hospitals, used older data, and/or did not examine PSIs or the current patient experience metrics used by CMS.

The research letter listed several possible explanations for the results:

  • The VA may have invested considerable efforts in quality improvement and care coordination over the past three decades;
  • The agency might have better documentation of comorbidities that are used in the risk adjustment, as a result of the VA’s unified electronic health record system and because patients have their health care mostly within one system;
  • While the mortality and readmission measures are thought to be indicative of quality, many of the Agency for Healthcare Research and Quality PSIs have been questioned due to the validity of the data—complications are coded inconsistently or data are not audited—and due to ascertainment bias.

The researchers pointed out that, when ascertainment bias occurs, quality measures may reflect the inverse of quality, where hospitals performing worse may be the ones providing higher-quality care.

  1. Blay E Jr, DeLancey JO, Hewitt DB, Chung JW, Bilimoria KY. Initial Public Reporting of Quality at Veterans Affairs vs. Non-Veterans Affairs Hospitals. JAMA Intern Med. 2017 Apr 17. doi: 10.1001/jamainternmed.2017.0605. [Epub ahead of print] PubMed PMID: 28418527.