Discussion

In this large, nationwide study of Medicare beneficiaries, we found no evidence that elective total hip or knee arthroplasty performed on superstition-associated calendar dates — Friday the 13th or during Mercury retrograde — was associated with increased revision risk, mortality, or changes in procedure volume. Hazard ratios for all outcomes were close to unity, and the results were consistent across joint type, revision type, and analytical approach.

While one might assume that superstition commonly takes a back seat in academic affairs, multiple studies have been conducted on the topic. For example, the “black cloud” phenomenon, referring to surgeons who work unusually harder or handle more complex cases, can still be found in the medical community [10]. Additionally, there exists a belief that uttering the word “quiet” increases the workload in a clinical setting, but no evidence was found to support this in three randomized controlled trials [11–13].

The majority of previously conducted studies have affirmed the findings presented here, indicating no association between perceiving Friday the 13th as a day of misfortune and clinical outcomes. In the field of orthopedics, only one study has thus far addressed this topic. Nardelli and colleagues investigated the pre- and postoperative Western Ontario and MacMaster Universities Osteoarthritis Index (WOMAC) score, and they reported no significant difference in patients undergoing total knee arthroplasty (TKA) on Friday the 13th compared to those operated on other days [14].

Also in other fields, research has shown that Friday the 13th has no influence on intraoperative blood loss and emergency frequency in a study involving 27,914 patients undergoing general, visceral, or vascular surgery [8]. Similarly, in the context of cataract surgeries, Faschinger et al. analyzed outcomes for 16,965 cases and found no statistically significant differences in the number of complications between surgeries performed on Fridays that fell on the 13th and those on other dates [15]. Therefore, the results presented here, based on a large registry database, can provide a sense of certainty and reduce stress for patients who may not feel comfortable or even experience anxiety about undergoing surgery on dates associated with superstitions. This, in turn, could positively influence the healing process [16].

Limitations

The Medicare data used in this study primarily consists of administrative claims records, which, like any such data, comes with inherent limitations when utilized for orthopedic outcomes research [17, 18]. Notably, we lack access to the detailed clinical records of these patients, and essential clinical measures (such as radiologic imaging findings) and other indicators (like blood chemistry and organ function tests) are not encompassed by the diagnosis or procedure codes in this dataset. Furthermore, important clinical factors such as frailty indices, precise body mass index (BMI) values (beyond the binary obesity diagnosis code), and measures of surgical complexity (e.g., operative time, surgical approach, implant type) were not available in the claims data. While these factors may influence outcomes, there is no reason to expect systematic differences between superstition-associated and control dates, and thus residual confounding is unlikely to bias the comparative findings. While individual records may have occasional errors of omission or commission, it is improbable that systematic errors are present across the millions of records within the entire Medicare system. One significant strength of this data lies in the wide range of facilities submitting these claims and the vast number of records processed by Medicare, providing a unique advantage. Moreover, the extensive information on patient characteristics and complications is of high quality due to its relevance for billing costs and its input by specialized professionals. Additionally, the use of Medicare claims data introduces structural considerations inherent to an insurance-based healthcare system. Healthcare access, utilization patterns, and diagnostic coding are influenced by reimbursement structures, which may introduce selection and information bias. Although such biases are unlikely to differ systematically across superstition-associated and control calendar dates, they may affect absolute risk estimates and limit the generalizability of our findings to universal or non-insurance-based healthcare systems.