Impact of summer environmental heat on colorectal surgical outcomes
Elizabeth Yates1, Keith Spangler2, Stanley Ashley1, Gregory Wellenius2, Louis L. Nguyen3
1Surgery, Brigham and Women's Hospital, Boston, Massachusetts, United States, 2Environmental Health, Boston University School of Public Health, Boston, Massachusetts, United States, 3Vascular and Endovascular Surgery, Brigham and Women's Hospital, Boston, Massachusetts, United States
Objective. To characterize the impact of high summer environmental heat on colorectal surgical outcomes.
Design. Retrospective cohort. We obtained surgical outcomes data from the New York State Inpatient Database (SID) and linked it to monthly averages of environmental temperature data obtained from North American Land Data Assimilation System (NLDAS) by month and zip code.
Setting. New York hospitals participating in the Healthcare Utilization Project (HCUP).
Patients. Patients who underwent colectomy or rectal resection in a New York hospital between May - September during the years 2016-2017.
Intervention(s). None.
Main Outcome Measure(s). In-hospital mortality.
Results. Among the 8523 patients in our cohort, 3964 (46.5%) patients were male, the average age was 63.8 years and the average Elixhauser score of 3.1. The cohort was predominantly White (5927, 69.5%) followed by Black (892, 10.5%) and Hispanic (752, 8.8%). Median length of stay was 6 days (0, 278). There were 300 (3.5%) patients who died before discharge and 1600 (18.8%) patients were discharged to an external care facility. Within 30 days, 804 (9.4%) patients were readmitted. Monthly temperature ranged from 55.5 F to 87.5 F with an average of 76.1 F during the study window. In both bivariable and multivariable analyses, high average monthly temperature was not associated with length of stay, 30-day readmission nor length of stay.
In bivariable analyses, an average monthly temperature higher than the 95th percentile (>86.4 F) trended toward a 54% increased odds of in-hospital mortality (p=0.056, 95% CI 0.962-2.355). In backward stepwise logistic regression controlling for patient factors (age, sex, race, comorbidities); community factors (median income, rurality); and surgical factors (emergency status, open vs laparoscopic), monthly temperature > 95th percentile was significantly associated with a 64.9% increased odds of in-hospital mortality (Table 1; p=0.48, 095% CI 0.981-2.659). In subgroup analyses, the same associations were seen in cohorts limited to open surgeries or emergent surgeries, but not in laparoscopic or scheduled operations.
Conclusions. Our study utilizing a novel approach to assess the impact environmental temperature on colorectal surgical outcomes suggests that exposure to severe environmental heat contributes to colorectal surgery mortality risk in the northeastern region of the United States (US). Sub-analyses suggest patients undergoing unscheduled operations (which are frequently performed via open approach) are not optimized and ultimately more vulnerable to extreme temperatures. With climate change driving up the frequency and intensity of heatwaves across the US, such work will become increasing important to understanding surgical outcomes.
Table 1. Backward Stepwise Logistic Regression for In-Hospital Mortality following Colectomy or Rectal Resection
| Odds Ratio | P-value | 95% Confidence Interval, Lower Bound | 95% Confident Interval, Upper Bound | |
| Age (10 year increase) | 1.154 | 0.004* | 1.047 | 1.275 |
| Elixhauser Score | 1.492 | <0.001* | 1.410 | 1.579 |
| Open Surgery | 5.469 | <0.001* | 3.402 | 9.400 |
| Emergent/Urgent Surgery | 3.760 | <0.001* | 2.827 | 5.062 |
| Race/Ethnicity | ||||
| White (referent) | -- | -- | -- | -- |
| Black | 1.335 | 0.119 | 0.921 | 1.906 |
| Hispanic | 0.763 | 0.316 | 0.437 | 1.262 |
| Asian or Pacific Islander | 0.578 | 0.298 | 0.172 | 1.439 |
| Other | 1.510 | 0.084 | 0.926 | 2.368 |
| Rurality | ||||
| Large metro (referent) | -- | -- | -- | -- |
| Small metro | 1.566 | 0.048* | 1.112 | 2.180 |
| Town (10-50k) | 1.007 | 0.980 | 0.587 | 1.647 |
| Rural (<10k) | 1.138 | 0.716 | 0.536 | 2.177 |
| Average Monthly Temperature > 95%ile | 1.649 | 0.048* | 0.981 | 2.659 |
*Significant at the p < 0.05 level
Back to 2022 Posters