There is an increasing emphasis on “rating” or “ranking” hospitals and doctors. Although much of the data is government-derived, this process is frequently either hospital-driven or media-driven. The hospitals with good results use the ratings and rankings in their marketing to patients. The media takes the rating data (which may be unranked) and uses them to rank hospitals and doctors, usually emphasizing the “good” hospitals and the “bad” hospitals. This is done with little or no concern for the accuracy of the data, the nuances of the data, or the significance of the differences. The parameters for ranking are variable, but the more dramatic, the better, from the media viewpoint. This is a process that we as physicians have to learn to accept but, at the same time, try to make as accurate and as appropriate as possible. Fortunately, the rating process has a beneficial side as it has provided a stimulus to improve treatment.

Doctors and hospitals have taken on somewhat adversarial roles in the present economic environment, and as shown in the paper by Kester et al., the “quality” of the hospital is rightly dependent on the physician’s out-of-hospital care. The health system (doctors and hospital) have to work together for improvements in quality. One of the issues with “rankings” and markers for “quality” is that the data have to be accurate, fair, and meaningful in order to be credible. The surgeon’s interest in high rankings for the hospital is obvious—patients will eventually seek out a different doctor if the surgeon’s hospital receives poor rankings. The CMS (Centers for Medicare & Medicaid Services) web site lists affiliated hospitals for each surgeon who accepts Medicare. Thus, it tends to encourage the surgeons and hospitals to work together. The hospital’s interest in the rankings is largely economic. A better ranking generally translates into lower costs, particularly for events designated as “never” events by the CMS, such as VTE (venous thromboembolism) in patients treated with total hip or knee arthroplasty. “Never” events are not reimbursed by Medicare.

This study evaluated the thirty-day VTE outcomes in a large data set involving more than 23,000 patients who underwent total or partial hip or knee replacement. A major strength of the study is that a majority of postoperative VTE events would be expected to take place in this time interval. Additionally, this article included post-discharge data, which have previously been lacking in a large study population. The study had trained analysts to evaluate the data, which probably substantially improved the quality and credibility of the data. Data derived from hospital data systems routinely have coding errors, such as the wrong procedure, wrong surgeon, and incomplete data.