To the Editor: As new research findings with clinical application are constantly being disseminated, psychiatrists and other physicians need knowledge about statistics in order to properly assess them. In a recent study of 277 internal-medicine residents, 95% agreed or strongly agreed that, to be an “intelligent reader of the literature,” it is necessary to know something about statistics (1). Unfortunately, when given a statistical knowledge test, this same group of residents answered on average only 8 out of 20 questions correctly. The Accreditation Council for Graduate Medical Education (ACGME) addresses this need in competency, requiring residents to “apply knowledge of statistical methods to the appraisal of clinical studies” (2). Importantly, this requirement is for all psychiatry residents, not just those preparing for a career involving research.

Given this requirement, the question becomes: How are we educating psychiatry residents to be statistically literate? In a pilot study, we surveyed general-psychiatry residency directors to develop a preliminary picture of the teaching of statistical literacy, defined as the ability to read and understand articles with statistical analysis, to non–research-track residents. An 18-item online survey was designed by the three authors and administered to 10 psychiatry residency directors selected by convenience sample. Participants were asked whether statistical literacy was taught at their program, and how important it was to teach this topic. Additional items included methods of instruction, recommendation on when statistical literacy should be taught, faculty expertise in statistics, and barriers to teaching this subject. Finally, the residency directors were questioned about their potential interest in various types of curricula that could be used to teach statistical literacy to general-psychiatry residents. Descriptive statistics were used to report response rates. All residency directors responded to the survey (N=10). Seven included statistical literacy in their residency program curriculum for non–research-track trainees. Five replied that it is very important to teach statistical literacy to psychiatry residents; four reported that it is moderately important; and one answered that it is minimally important. No residency directors responded that it is not important to teach residents about statistical literacy. Of those who provide education on this topic, multiple methods were used, with the journal club attended by a faculty member with formal statistical training most common.

Formal curricula on statistical literacy were limited: four programs reported a short class in statistical literacy, and one provided a semester-long course. Of those programs that taught statistical literacy, three began educating residents during PGY I; two initiated teaching during PGY II; and two started in PGY III. Seven of the 10 residency directors believed that statistical literacy should first be taught in medical school. Nine of the programs reported that their faculty included individuals with formal training in statistics, and five reported that their faculty provided statistical expertise to other departments.

Eight directors endorsed competing priorities as a barrier to teaching statistical literacy. Other barriers included lack of interest among residents (N=6), lack of faculty expertise (N=2), and budgetary restraints (N=1). Eight residency directors expressed interest in a portable curriculum addressing statistical literacy. Six responded that they would like to see a special column in an education or general-psychiatry journal devoted to this topic. Of this predominately university-based sample, most respondents stated that it was important to teach statistical literacy to non–research-track residents, yet few programs offered a formal curriculum on this topic. In fact, only one program offered a semester-long course. The journal club was the most common method used for teaching statistical literacy in this group. Unfortunately, little data support this method as an efficacious educational tool. The most recent review of the journal club in postgraduate medical education indicated improvement in residents' knowledge of biostatistics, but long-term retention of information is questionable (3, 4). Seven of the residency directors in our survey stated that statistical literacy should first be taught during medical school, the earliest time-period for possible response. In the development of a curriculum on “physician numeracy,” which Rao and Kanter define as “understanding the statistical aspects of, and the terminology associated with, the design, analysis, and results of original research,” the authors state that one major challenge they faced was the fact that the topic was “easily forgotten if not reinforced” (5). Repeated exposure to statistical concepts and terminology should improve retention, suggesting that even if statistical literacy is taught in medical schools, psychiatry residency programs need to reinforce it. Survey respondents declared interest in both a portable curriculum and a regularly-scheduled educational column devoted to this subject. Perhaps this is a way in which our academic community could address some of the issues related to the teaching of statistical literacy that this preliminary picture revealed.

Manuscripts authored by an editor of Academic Psychiatry or a member of its editorial or advisory board undergo the same editorial review process, including blinded peer-review, applied to all manuscripts. Also, the editor is recused from any editorial decision-making.