Complications and Month of Surgery: Does Scheduling Make a Difference?

Introduction

Adverse events (AEs) related to surgical procedures are a significant cause of morbidity and mortality, and account for up to 40% of hospital complications. ¹ Such events are becoming increasingly important for measuring health care quality, and research has focused on potentially modifiable influences in an effort to improve outcomes. Studies are often designed to evaluate either patient-related factors or surgeon-related factors associated with increases in AEs.

A systematic review of 30 articles reported 9 significant patient factors related to surgical outcomes. ² Among these, age, body mass index (BMI), and American Society of Anesthesiologists classification were the most significantly associated factors. Other researchers have described higher rates of surgical-site infections, hospital-acquired infections, adverse drug reactions, and lengths of stays among patients with recent hospitalizations prior to surgery.^3–5

Patients who meet their insurance deductibles are financially incentivized to proceed with elective procedures, as these patients will have reduced personal costs. Nevertheless, it is possible that such patients could have a higher burden of health concerns that resulted in them meeting their deductibles that year. Thus, it was posited that more complicated cases, or patients whom have been recently hospitalized might be more likely to undergo surgery at the end of the “insurance year,” compared to other times of the year.

A commonly studied surgeon factor is the “July effect,” corresponding to the beginning of a new academic year at teaching hospitals. Findings have been conflicted on if there are differences in outcomes and complications during that time of the year. Englesbe and colleagues found an 18% higher postoperative morbidity rate and a 41% high mortality rate in July–August, compared to April–June. ⁶ Likewise, another study found higher morbidity and mortality in July for patients after surgery for spinal metastases. ⁷ However, Highstead et al. found no differences in complications in July–August, compared to April–May for more than 3000 patients admitted to a Level 1 trauma service. ⁸

The objective of this study was to evaluate if there were differences in perioperative morbidity associated with elective surgery based on the month in which the surgery was performed. It was hypothesized that surgical complication rates would be higher at the end of the calendar year (November/December) than in other months, as the latter months of the year often correspond with insurance deductible year-endings. Another goal was to evaluate if the end of June (another common insurance deductible year-end) or July (start month of resident trainees) would be related to an increased likelihood of perioperative morbidity. ⁹

Materials and Methods

Results

A total of 46,590 procedures were performed during the 2-year study period. Twenty-thousand subjects were randomly selected (833 per month). After exclusions, 19,872 subjects were included in the analysis. Baseline characteristics (shown in Table 1) did not differ among the months. The mean age was 57 and the mean body mass index was 31. There were more female than male subjects (62.6% versus 37.4%, respectively). The majority of patients were white (84%). Regarding payer status, 51.4% had commercial insurance and 46.9% had Medicare/Medicaid. Procedures were most commonly performed by General Surgery (25.3%), Orthopedics (23.2%), and Gynecology (15.2%). Outpatient procedures comprised 53.6% of the surgeries, with the remaining 46.4% classified as inpatient surgeries. Most procedures were performed using general anesthesia (85.5%). At least 1 medical comorbidity was noted in 64.9% of surgical subjects. The most common comorbidity was hypertension (51.9%), followed by pulmonary disease (21.4%) and diabetes mellitus (19.1%; Table 1).

Complications were associated with 17.7% of the surgeries. The months with the highest number of complications were June (n = 368) and November (n = 374; Table 3). The most common complications reported were reoperation (10%), readmission (5.9%), and blood transfusion (2.5%; Table 2). Logistic regression analysis showed the odds of having any complication was 1.4 times greater for June (odds ratio [OR] = 1.38; 95% confidence interval [CI]: 1.22–1.55; p < 0.001) and November (OR = 1.41; 95% CI: 1.25–1.59; p < 0.001) than other months (Table 4). When controlling for potential confounders such as age, sex, inpatient/outpatient, and presence of resident involvement, a significant relationship was still observed (Table 5). There were also a higher number of subjects with more than 3 complications during these 2 months: n = 39 in June (p ≤ 0.001), and n = 49 (p ≤ 0.001) in November (Table 4). However, December was associated with a decrease in complications (n = 242; OR = 0.78; 95% CI: 0.68–0.90; p = 0.001), compared with other months.

Regarding trainee involvement, complications recorded in the month of July (n = 309) were not significantly higher than other months of the year (OR = 1.07; 95% CI: 0.94–1.22; p = 0.322). Furthermore, trainee involvement in care was not associated with increased complications (OR = 1.05; 95% CI: 0.97–1.13; p = 0.240). When analyzing by primary payer status, significantly higher rates of complications were noted in Medicare/Medicaid patients, compared with patients who had commercial insurance (OR = 1.97; 95% CI: 1.83–2.13; p = < 0.001; data not shown).

Discussion

This retrospective study evaluated differences in perioperative morbidity and mortality based on months in which surgeries were performed. Patients who underwent surgery within this hospital system from 2015 to 2016 were at significantly increased risk of complications during the months of June and November, with the most common complications being reoperation, readmission, and blood transfusion. Patients who used Medicare/Medicaid as their primary insurance were also more likely to experience perioperative complications, compared to patients with private insurance. Reassuringly, trainee involvement did not increase the risk of complications at any time of the year.

While there is literature on the July effect,^6–9 there is a paucity of data on surgical outcomes and their temporal relationships to end-of-insurance terms. Patients who meet their insurance deductibles are likely more eager to proceed with an elective procedure, as they will have little-to-no out-of-pocket cost. Many physician practices notice a surge of patients seeking care at the end of these terms, compared to other times of the year. Patients who have met their deductibles likely utilize more health care resources because these patients have more health problems or have had more procedures performed during that year. As most insurance terms align with the calendar or academic year, the current results might support the notion that risk of complication increases near the end of the insurance terms, such as November and June. While it was surprising that this relationship did not continue in the month of December, it is possible that the pending holiday season could preclude patients from seeking more-invasive procedures in the last month of the year. Nevertheless, the current authors believe collaboration with insurance companies may be warranted to lessen the incentives to seek surgical interventions, particularly if this behavior might contribute to higher complication rates.

Several studies have focused on if trainee involvement in surgical cases affects the frequency or severity of postoperative complications. Many of these studies focused on specific surgeries or surgical specialties, such as hysterectomy, bariatric surgery, and emergency general surgery. In these reports, trainee involvement was associated with more-frequent and -severe adverse surgical outcomes.^11–13 However, studies that have taken broader views of academic and nonacademic centers, and evaluated a range of specialties, did not show significant differences with trainee involvement. ¹⁴ The current study included a variety of surgeries in a high-volume, community-based teaching-hospital system with significant trainee involvement. The current study showed that trainee involvement alone was not associated with increased risks of postoperative complications, similar to studies that evaluated a range of surgical procedures. This might provide reassurance for patients seeking care at an academic center, particularly during the months of July and August.

Physicians strive to provide the same high-quality care for patients regardless of the source of the primary payer for such care. However, one cannot ignore how primary payer status may predict perioperative risks. Xu et al.'s findings supported this notion and postulated that Medicaid insurance status should be viewed as a perioperative risk factor for complications. ¹⁵ As Xu et al. rightfully note, confounders such as race, ethnicity, socioeconomic status, BMI, postoperative treatment, and pain management must be considered when evaluating this risk, and improvements can be made by expanding provider education about the topic. ¹⁵ The current study attempted to control for many of these confounding variables; however, the significant disparity of postoperative complications persisted, consistent with the reported literature.^15,16 Steps to mitigate this discrepancy may include more aggressive preoperative health optimization, improved patient education, and closer follow-up after surgery.

Strengths of the current study included analysis of a large sample of patients across multiple surgical specialties. All the procedures were performed within a large community-based teaching-hospital system. This enabled analysis of populations in both Medicaid/Medicare and commercial insurance systems.

Limitations of the current study included those inherent to retrospective cohort studies, such as information bias or mis-classification. With the retrospective use of medical records, the current authors had to rely on other clinicians for accurate recordkeeping. Despite manual chart audits, the current authors' reported readmission and reoperation rates might be inaccurate due to an inability to delineate subjects who might have returned to the hospital or operating room during the subsequent 90 days for an unrelated diagnosis or indication. While elective surgeries performed in main operating room suites were evaluated, selections were not only made for inpatient procedures and the current authors could not comment specifically on types or invasive nature of surgeries. The current authors were also only able to identify subsequent encounters within the studied hospital system and thus were unable to determine encounters at other hospitals. Most of the subjects in this current study were white, and the results might not be representative of other populations. Additionally, the current authors acknowledge the inability to truly measure the degree and level of trainee participation in the surgeries. Finally, it was not possible to track how many procedures each individual might have underwent in the preceding months, due to limitations in the database; thus, the current authors could specifically associate having had recent surgery with the higher rates of complications.

Conclusions

Large multicenter studies are needed to investigate trends in surgical complications further and confirm the associations noted here. Such investigations into the seasonal variations ⁶ as well as the day/time/week variations in perioperative mortality ¹⁷ have buffered the evidence, although more is needed. Ultimately, proper patient selection and appropriate timing for elective procedures might have a positive effect against the adverse outcomes frequently identified. Ultimately, reducing perioperative morbidity has great benefit in improving quality of care and lowering health care costs. Greater scrutiny regarding end-of-year financial incentives may be warranted to ultimately improve patient outcomes and reduce the burdens of these complications.