To the best of our knowledge, this is the first study evaluating whether the arrival of new residents leads to seasonal variation in the quality of requests for liaison psychiatric consultation. Our results show higher rates of inaccuracy between the requesters and psychiatric consultants during the period of arrival of new residents (April-May), as well as during and after the year-end holidays, with conclusive evidence of a direct causal association with the requester being a resident.
Seasonality effects have been demonstrated to affect general healthcare systems for many reasons. For example, the influx of new trainees, which usually occurs in the summer is often associated with reduced quality of care and higher mortality, the so-called "July Effect." [12, 13]. However, during these same periods, patients are less likely to undergo elective surgeries, and surgeons are more likely to be on vacation. As a result, the number of emergency surgeries tends to proportionally increase, with an accompanied rise in risks to patients [13].
The increase in the rates of psychiatric misdiagnoses in general inpatients at the beginning of the academic year is compatible with results from previous studies which associated the presence of new trainees with a rise in medical errors [12]. Of importance, clinical handoffs are considered essentially dangerous contextually, given the possibility of errors in the process of information transfer, as well as being associated with decreased adherence to treatment possibly because this change leads to discomfort in some patients [5, 14, 15]. As a result, the participation of a less experienced team of practitioners in an already complicated process is of particular concern and calls for caution.
There are many possible explanations to the deterioration in the quality of the requests during the year-end holidays and the subsequent summer vacation. First, during year-end holidays, hospital routines are significantly affected, with a reduction in the number of active staff members and, as a consequence, a change in teams that often worked together. This often results in a change in routines, and errors become more frequent. A second possibility, particularly pertinent in the context of residents is that with fewer faculty members, more unsupervised decisions are made, ultimately increasing error rates. Lastly, the period following the year-end holidays has been associated with a worsening in a number of mental health conditions, perhaps due to psychological stress [16]. As a consequence of the change in the usual pattern of mental health conditions, diagnoses become more complex, and error rates increase. In practice, it is likely that a combination of all these factors might lead to an increase in inaccuracy rates, although our study did not have the information to draw any conclusions about the underlying causes.
Despite filling an important gap in the literature, our study does have limitations. First, our sample only represents the experience of a single, tertiary, academic hospital; thus results could vary in other contexts. However, our institution is not atypical when compared to most educational services and therefore we believe that result fluctuations should not affect our main conclusions. Second, although our analysis led to a causal association between requesters being residents and an increase in inaccuracy rates, the number of requests by non-residents was relatively small. Nevertheless, given that the results were statistically significant, we believe that an evaluation of institutions with a larger number of requests by non-residents would likely increase rather than decrease the effect we have reported.