Commentary by Danise Schiliro-Chuang, NYU Chief Resident
Welcome to this week’s edition of ShortCuts. Hope that everyone has enjoyed the nice spring weekend weather.
Let’s begin with some news in the world of medical informatics. In an article published in the April 17 edition of NEJM and that grabbed the attention of the New York Times, authors Mandl and Kohane discuss the pros and cons of personally controlled health records (PCHRs). Two large corporations, Microsoft and Google, have begun to offer patients Web-based personal health records. Under such a system, patients give consent to fill their own PCHR with data from multiple health care sites. Patients can also authorize access to information to other individuals, such as health care providers, health care proxies, and researchers, and can sign up for notification of open research studies. The advantages of such a medical record system might seem obvious to us as practicing clinicians. For one, patients and their providers would have access to an integrated medical record. If your patient is hospitalized at an outside institution and has an emergency room visit to yet another outside facility, you and your patient would not have to go through the often cumbersome process of recovering the data from these encounters. Such a system would also empower patients by keeping them in the loop with regards to the tests they undergo and medications they receive. Finally, patient enrollment in research studies might increase given the easier access to patient data and the easier access of patients to the studies being conducted. However, as Mandl and Kohane point out, this system does not come without risks to both patients and academic medical centers. They note that while consumer-driven data aggregation might empower the patient, companies like Microsoft and Google are not bound by the privacy restrictions of the Health Insurance Portability and Accountability Act (HIPAA). While HIPAA regulations define what constitutes a de-identified data set and what the penalties are for infractions in their use and management, PCHR companies are not bound by these regulations. In addition, academic medical centers might find themselves competing with PCHR vendors for the participation of their own patients in research studies. Mandl and Kohane propose that PCHR companies should all clearly define whether and how they will de-identify data and that protection for research subjects needs to be at least partially regulated by the government by extending HIPAA to include online patient record hosts. Stay tuned. I’m sure there will be a lot more on this hot topic in the very near future.
A meta-analysis published in this week’s Annals of Internal Medicine examines the use of topical or intraluminal antibiotics to prevent hemodialysis catheter-related bloodstream infections, a cause of substantial morbidity and mortality. The authors, James et al, looked at 15 randomized controlled trials comprising 1294 patients undergoing hemodialysis via a catheter who received topical or intraluminal antibiotics for infection prophylaxis. The meta-analysis found that both types of antibiotics reduced the rate of bacteremia as well as the need for catheter removal secondary to complications. While limited by a small number of studies, few of which were blinded, short duration of follow-up, publication bias, and a paucity of data on the adverse effect of antibiotic resistance, the positive findings of this study warrant additional research with longer follow-up periods and harder clinical end-points.
Finally, in this week’s issue of JAMA, Ross et al explore the practice of ghostwriting in publications related to the drug, Rofecoxib (Vioxx). As we know, Vioxx was a best-selling drug for the pharmaceutical company, Merck, before it was taken off the market in 2004 after evidence linked it to myocardial infarctions. Following this, thousands of lawsuits were filed against Merck. Dr. Ross, who served as a consultant to plaintiffs’ lawyers in several such lawsuits, was given access to internal Merck e-mails and documents regarding 96 journal publications. Stemming from this, Ross and his colleagues published a case study of these industry documents. This case study reveals data showing that, in many cases, articles related to Vioxx were authored by Merck employees but attributed first authorship to external, academically affiliated investigators who did not always disclose financial support from Merck, although financial support of the study was nearly always provided. While both Merck as well as several first authors named in the article refute the assertion that they did not contribute substantially to the final publications, this case study raises many questions about the validity of drug industry published research. An accompanying editorial calls for editors of medical journals to require authors to disclose their specific contributions to the articles.