The appropriateness of 30-day mortality as a quality metric in colorectal cancer surgery

doi:10.1016/j.amjsurg.2017.04.018

The American Journal of Surgery

Volume 215, Issue 1, January 2018, Pages 66-70

https://doi.org/10.1016/j.amjsurg.2017.04.018 Get rights and content

Abstract

Background

Our study compares 30-day vs. 90-day mortality following colorectal cancer surgery (CRS), and examines hospital performance ranking based on this assessment.

Methods

Mortality rates were compared between 30 vs. 90 days following CRS for patients with stage I-III colorectal cancers from the National Cancer Database (2004–2012). Risk-adjusted hierarchical regression models evaluated hospital performance based on mortality. Hospitals were ranked into top (10%), middle (80%), and lowest (10%) performance groups.

Results

Among 185,464 patients, 90-day mortality was nearly double the 30-day mortality (4.4% vs. 2.5%). Following risk adjustment 176 hospitals changed performance ranking: 39% in the top 30-day mortality group changed ranking to the middle group; 37% of hospitals in the lowest 30-day group changed ranking to the middle 90-day group.

Conclusions

Evaluation of hospital performance based on 30-day mortality is associated with misclassification for 15% of hospitals. Ninety-day mortality may be a better quality metric in oncologic CRS.

Introduction

Colorectal cancer is the third most common malignancy affecting both males and females in the United States.¹ It is estimated that more than 134,400 patients will be diagnosed with the disease in 2016.² Surgical resection represents the mainstay of treatment for patients with localized colorectal tumors.³

Colorectal surgery is an ideal target for quality improvement initiatives because they are some of the most frequent procedures performed in the inpatient settings and they carry a significant morbidity with a mortality ranging from 4 to 10%.4, 5 Surgical mortality is considered the most important quality metric of surgical care generally and in colorectal surgery. It is currently used as a yardstick to measure quality of performance of hospitals and surgeons in the United States. Estimate of surgical mortality is a realistic measure for operative risk and helps informs patients and clinicians on treatment decisions. Therefore, accurate representation of colorectal surgical mortality is crucial and of great interest to patients, surgeons and non-surgeon providers, policy makers, and payers.

Traditionally, mortality within 30 days (30-day mortality) from surgery has been used to represent surgical mortality. The American College of Surgeons National Surgical Quality Improvement Program (ACS-NSQIP) and ACS-NSQIP Surgical Risk Calculator define surgical mortality as death within 30 days of surgery.6, 7 However, published data have suggested that incidences of surgical mortality continue beyond the first 30 postoperative days for cancer patients undergoing esophagectomy and pancreatectomy.8, 9, 10 In colorectal surgery, limited data exist, which suggest a significant number of surgical deaths occur beyond the first 30 days after surgery.5, 11 In a single-institution study of 186 patients undergoing colorectal surgery, 30-day mortality was 4.3%, which increased to 9.1% by 90 days from surgery.¹¹ These findings impugn the appropriateness of 30-day mortality in capturing the true incidence of surgical mortality and its accuracy as a critical quality metric. Therefore, we sought to compare incidence of 30-day vs. 90-day mortality after colorectal cancer surgery and examine whether hospital performance ranking changes based on 30-day vs. 90-day mortality.

Section snippets

Materials and methods

The National Cancer Data Base (NCDB) is a joint program of the American Cancer Society and the Commission on Cancer (CoC) of the American College of Surgeons. The NCDB is a nationwide, facility-based, comprehensive dataset, which captures >70% of some newly diagnosed malignancies in the United States. Established in 1989, it contains >29 million cancer cases from >1500 CoC-accredited cancer programs from all states, Puerto Rico, and the District of Colombia. It captures >85% of all new cancer

Results

A total of 185,464 patients who underwent major colorectal surgery were identified: 56% (n = 104,041) had colon cancer, 29% (n = 53,552) rectal cancer, and 15% (n = 27,871) rectosigmoid cancer. Of all patients, the rate of 30-day mortality was 2.6% (n = 4761), which nearly doubled within 90 days from surgery to 4.4% (n = 8098); 90-day mortality rates remained similarly higher regardless of cancer site (Fig. 1).

Table 1 details patient demographic, clinicopathologic, and treatment characteristics

Discussion

This large study of 185,464 colorectal cancer patients undergoing surgery explored the appropriateness of 30-day mortality as a quality metric of colorectal surgical care in the United States. Thirty-day mortality appears to significantly underestimate incidence of death after colorectal surgery, with a significant number of deaths occurring beyond 30 days. The rate of mortality by 90 days from surgery was nearly double the rate of 30-day mortality. Evaluation of hospital performance based on

Conclusions

This large study explored the appropriateness of 30-day mortality as a quality metric in colorectal surgery for patients with cancer. Thirty-day mortality appears to underestimate incidence of death after colorectal surgery, with nearly doubling of the mortality rate by 90 days postoperatively. This finding may inform providers and patients about the true risk for operative mortality and facilitate the surgical risk-benefit discussion. Evaluation of hospital performance based on 30-day

Disclosure

A portion of the data were presented as a plenary presentation at the Society for Surgery of the Alimentary Tract 57th Annual Meeting, May 21–24, 2016, San Diego, California.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

References (20)

B.A. Yerokun et al.
Does conversion in laparoscopic colectomy portend an inferior oncologic Outcome? Results from 104,400 patients
J Gastrointest Surg
(2016)
R.L. Siegel et al.
Cancer statistics
CA Cancer J Clin
(2016)
American Cancer Society Cancer Facts & Figures
(2016)
National Comprehensive Cancer Network. Clinical Practice Guidelines in Oncology...
B.E. Byrne et al.
Population-based cohort study comparing 30- and 90-day institutional mortality rates after colorectal surgery
Br J Surg
(1810-1817)
American College of Surgeons. Cancer Quality Improvement Program. Available at:...
American College of Surgeons National National Surgical Quality Improvement Program. Surgical Risk Calculator....
J.D. Birkmeyer et al.
Hospital volume and surgical mortality in the United States
N Engl J Med
(2002)
H. In et al.
Doubling of 30-day mortality by 90 Days after esophagectomy: a critical measure of outcomes for quality improvement
Ann Surg
(2016)
R.S. Swanson et al.
The 90-day mortality after pancreatectomy for cancer is double the 30-day mortality: more than 20,000 resections from the national cancer data base
Ann Surg Oncol
(2014)

There are more references available in the full text version of this article.

Cited by (27)

Five-meter walk test before transcatheter aortic valve replacement and 1-year noncardiac mortality
2022, JTCVS Open
The purpose of this study is to assess whether the 5-m walk test is associated with 1-year mortality after transcatheter aortic valve replacement.
Included in the analysis were 304 patients who received the 5-m walk test and underwent transcatheter aortic valve replacement from September 2012 to March 2019. They were classified into 3 groups based on their test score: ≤7, >7, and unable to walk. Preprocedure characteristics, postprocedure outcomes, and follow-up outcomes were compared between the groups.
For the 5-m walk test, 145 had a score ≤7 (Group N), 111 had a score >7 (Group S), and 48 were unable to walk (Group I). Average age in years was 80.2 ± 8.7 years in Group N, 81.2 ± 9.4 years in Group S, and 79.4 ± 9.2 in Group I (P = .23). The aortic valve mean gradient at discharge was 9.5 ± 4.1 mm Hg in Group N, 10.4 ± 5.5 mm Hg in Group S, and 8.2 ± 4.2 mm Hg in Group I (P = .05). The discharge survival was 97.2% in Group N, 96.4% in Group S, and 95.8% in Group I (P = .76). One-year survival was 92.8% in Group N, 84.1% in Group S, and 75% in Group I (P < .01) after adjusting for preprocedure characteristics. Noncardiac death was 5.1% in Group N, 13.1% in Group S, and 22.7% in Group I (P = .03). This indicates that the 5-m walk test was a risk factor for 1-year mortality. More specifically, a poor 5-m walk test score was associated with 1-year noncardiac mortality.
The 5-m walk test score before transcatheter aortic valve replacement was associated with 1-year mortality, especially noncardiac mortality. It may help identify patients at high risk for 1-year mortality.
Widening volume and persistent outcome disparity in valve operations: New York statewide analysis, 2005-2016
2022, Journal of Thoracic and Cardiovascular Surgery
Citation Excerpt :
Future studies on this relationship should analyze larger regional databases before generalizing to other states. Lastly, the mortality rates presented in the study are only based on 30-day mortalities, and other important longer-term outcomes such as readmissions and reoperations could not be evaluated.22,23 In NYS, over the last decade, highest-volume hospitals increased their market share for valve operations, and their RAMR decreased at a greater rate than that of lowest-volume hospitals.
Volume concentration of complex noncardiac operations to high-volume centers has been observed, but whether this is also occurring in cardiac surgery is unknown. We examined the relationship between volume concentration and mortality rates for valve surgery and coronary artery bypass grafting (CABG) between 2005 and 2016 in New York State.
We analyzed publicly available, hospital-level case volume and risk-adjusted mortality rates (RAMRs) from 2005 to 2016 for isolated CABG and isolated or concomitant valve operations performed in New York. We identified hospitals in the top- and bottom-volume quartiles for each procedure type and compared changes in percent market share and outcomes. Bivariate and univariate longitudinal analysis was used to evaluate the statistical significance of the temporal trend.
Among 36 centers, percent market share of the top-volume quartile increased for valve cases from 54.4% to 59.4%, whereas CABG share increased from 41.4% to 44.3%. No significant changes were noted in market share for the bottom quartile. The top-volume quartile demonstrated significant trends in improving outcomes over the study period for both valve procedures (RAMR: −0.261%/year, P < .001) and CABG (RAMR: −0.071%/year, P = .018). No significant trends were noted in the bottom quartile for either procedure.
In New York, over the last decade, highest-volume hospitals increased their market share for valve operations while maintaining lower mortality rates than lowest-volume hospitals. Valve volume is regionalizing in the setting of a persistent outcome gap between the highest- and lowest-volume hospitals, suggesting that volume-based referrals for specialized cardiac procedures may improve surgical mortality.
Failure to rescue after reoperation for major complications of elective and emergency colorectal surgery: A population-based multicenter cohort study
2022, Surgery (United States)
Citation Excerpt :
In addition, this study also aimed to identify risk factors for FTR, defined as mortality for any cause within 90 days after a reoperation after elective or emergency colorectal surgery. Instead of 30-day mortality, 90-day mortality was chosen because nearly half of the mortality occurs between 30 and 90 days, thus 90-day mortality serves as a better quality measure.22–25 Our hypothesis was that reoperation and FTR rates, as well as the spectrum of complications, differ between elective and emergency colorectal surgery.
As surgical complications inevitably occur, minimizing the failure-to-rescue rate is of paramount interest. Most of the failure-to-rescue research in colorectal surgery has previously focused on elective surgery and anastomotic dehiscence. The aim of this study was to characterize and compare the major postoperative complications demanding reoperation after elective versus emergency colorectal surgery, and to the identify risk factors for failure-to-rescue.
In this population-based retrospective multicenter cohort study, adult patients undergoing a reoperation for colorectal surgery complication between 2006 and 2017 in 10 hospitals were included. The data were manually extracted. Failure-to-rescue was defined as 90-day mortality after the reoperation.
In total, 14,290 patients underwent index colorectal resection, of which 862 (5.8%) underwent emergency reoperation within 30 days (438 [4.3%] after elective, 424 [10.4%] after emergency index operation, P < .001). The failure-to-rescue overall rate was 17.4% (8.0% after elective vs 27.1% after emergency index operation, P < .001). The 4 most common complications were anastomotic dehiscence (36.6%, 316 patients), fascial rupture (23.5%, 203 patients), intra-abdominal bleeding (15.3%, 131 patients), and bowel obstruction (10.2%, 88 patients). The majority (640 patients, 74.2%) had 1 of these complications; 261 patients (30.3%) had multiple complications. In multivariable analyses, the only possibly preventable independent risk factor for failure-to-rescue was perioperative organ failure at the time of reoperation.
Major complications requiring reoperation occur more than twice as often after emergency surgery and have a higher failure-to-rescue rate of >3× compared with elective surgery. The 4 most common complication types constitute three-fourths of the complications, providing a target for quality improvement.
Risk factors associated with postoperative morbidity and mortality in D2 radical gastrectomy for gastric cancer
2022, Revista de Gastroenterologia de Mexico
La cirugía es el tratamiento principal para el cáncer gástrico. La gastrectomía radical D2 está asociada a morbilidad y mortalidad postoperatoria variable a nivel mundial. El objetivo del estudio fue identificar los factores de riesgo asociados con morbilidad y mortalidad postoperatoria de la gastrectomía radical D2 con intención curativa por cáncer gástrico.
En este estudio de serie de casos retrospectivo, se revisó las historias clínicas de los pacientes con cáncer gástrico operados de gastrectomía radical D2 desde enero del 2014 a diciembre del 2018. Se realizó un análisis univariado y multivariado para identificar los factores de riesgo relacionados con la morbilidad y mortalidad postoperatoria a los 90 días.
En 691 pacientes, el porcentaje de morbilidad y mortalidad postoperatoria fue del 23.3% y del 3.3%, respectivamente. En el análisis multivariado, la edad ≥ 70 años (OR = 1.85, IC 95%: 1.25-2.76), ASA III-IV (OR = 2.06, IC 95%: 1.28-3.34), la realización de gastrectomía total (OR = 1.96, IC 95%:1.19-3.23) y pancreatoesplenectomía (OR = 5.41, IC 95%:1.42-20.61) se asociaron con mayor morbilidad postoperatoria. Por otro lado, la edad ≥70 años (OR = 4.92, IC 95%:1.78-13.65), un menor IMC (OR = 0.81, IC 95%:0.71-0.92) e hipoalbuminemia (OR = 0.91, IC 95%:0.85-0.98) se asociaron a una mayor mortalidad en la gastrectomía radical D2 distal y total.
La gastrectomía radical D2 por cáncer gástrico es un tratamiento seguro con una baja morbilidad y mortalidad postoperatoria. La edad ≥ 70 años, ASA III-IV, la gastrectomía total y la realización de pancreatoesplenectomía fueron factores asociados con una mayor tasa de complicaciones. La edad ≥ 70 años, un menor IMC e hipoalbuminemia son predictores de mortalidad en la gastrectomía radical distal y total.
Surgery is the main treatment for gastric cancer. D2 radical gastrectomy is associated with a variable postoperative morbidity and mortality rate worldwide. The aim of the present study was to identify the risk factors associated with the postoperative morbidity and mortality of D2 radical gastrectomy, with curative intent, for gastric cancer.
A retrospective case series was conducted, in which the medical records were reviewed of patients with gastric cancer that underwent D2 radical gastrectomy, within the time frame of January 2014 and December 2018. Univariate and multivariate analyses were carried out to identify the risk factors related to postoperative morbidity and mortality within 90 days.
The percentages of postoperative morbidity and mortality in 691 patients were 23.3% and 3.3%, respectively. In the multivariate analysis, age ≥ 70 years (OR = 1.85, 95% CI: 1.25-2.76), ASA III-IV (OR = 2.06, 95% CI: 1.28-3.34), total gastrectomy (OR = 1.96, 95% CI:1.19-3.23), and pancreatosplenectomy (OR = 5.41, 95% CI: 1.42-20.61) were associated with greater postoperative morbidity, and age ≥ 70 years (OR = 4.92, 95% CI:1.78-13.65), lower BMI (OR = 0.81, 95% CI: 0.71-0.92), and hypoalbuminemia (OR = 0.91, 95% CI: 0.85-0.98) were associated with greater mortality in distal and total D2 radical gastrectomy.
D2 radical gastrectomy for gastric cancer was shown to be a safe treatment, with low postoperative morbidity and mortality rates. Age ≥ 70 years, ASA III-IV, total gastrectomy, and pancreatosplenectomy were factors associated with a higher complication rate. Age ≥ 70 years, lower BMI, and hypoalbuminemia were mortality predictors in distal and total radical gastrectomy.
The Association Between Institutional Mortality After Coronary Artery Bypass Grafting at One Year and Mortality Rates at 30 Days
2022, Journal of Cardiothoracic and Vascular Anesthesia
Citation Excerpt :
As mentioned previously, these findings are not unique to cardiac surgery. Approximately 15% of colorectal surgery programs are misclassified as low-performing centers due to poor 30-day metrics; while their mid- to long-term data are significantly better.12 In 2014, McMillan et al. noted that in their thoracic surgery patients, more than half of the deaths after lung and esophageal resection occurred postoperatively between 30 and 90 days.13
To assess the association between the common quality metric of 30-day mortality and mortality at 60 days, 90 days, and one year after coronary artery bypass grafting.
A retrospective cohort study, with multivariate logistic regression to assess association among mortality outcomes.
Hospitals participating in Medicare and reporting data within the Centers for Medicare and Medicaid Services Limited Data Set between April 1, 2016, and March 31, 2017.
A total of 37,036 patients undergoing surgery at 394 hospitals.
None.
Mortality rates were 1.0%-to-3.1% for the top and bottom quartile of hospitals at 30 days. At one year, the top 25th percentile of hospitals had mortality rates averaging 3.9%; while hospitals below the 75th percentile had mortality rates averaging 7.6%. Twenty-three percent of hospitals in the top quartile at 30 days were no longer in the top quartile at 60 days. At one year, only 48% of hospitals that were in the top quartile at 30 days remained in the top quartile. The correlation between mortality rates at 30 days and the reported points was assessed using Spearman's rho. The R value between mortality at 30 days and mortality at one year was 0.53, which improved to 0.7 and 0.76 at 60 and 90 days.
Mortality at 30 days correlated poorly with mortality at one year. Hospitals that were high- or low-performing at 30 days frequently were no longer within the same performance group at one year.
Perioperative mortality as a meaningful indicator: Challenges and solutions for measurement, interpretation, and health system improvement
2020, Anaesthesia Critical Care and Pain Medicine
Citation Excerpt :
However, in high-resource settings, improvements in postoperative care have prolonged the interval between operation and postoperative death. In these high resource settings, the 90-day POMR is increasingly more discriminative and meaningful than either 30-day or inpatient POMRs [30,31]. While capture of 90-day POMR is not feasible in most low- or middle-income countrit (LMIC) settings, increasing digitisation and uptake or mobile phone technology may make this a realistic target in the future.
Expanding global access to safe surgical and anaesthesia care is crucial to meet the health targets of the Sustainable Development Goals (SDGs). As global surgical volume increases, improving safety throughout the patient care pathway is a public health priority. At present, an estimated 4.2 million individuals die within 30 days of surgery each year, and many of these deaths are preventable. Important considerations for the collection and reporting of perioperative mortality data have been identified in the literature, but consensus has not been established on the best methodology for the quantification of excess surgical mortality at a hospital or health system level. In this narrative review, we address challenges in the use of perioperative mortality rates (POMR) for improving patient safety. First, we discuss controversies in the use of POMR as a health system indicator and suggest advantages for using a “basket” of procedure-specific mortality rates as an adjunct to gross POMR. We offer then solutions to challenges in the collection and reporting of POMR data, and propose interventions for improving care in the preoperative, operative, and postoperative periods. Finally, we discuss how health systems leaders and frontline clinicians can integrate surgical safety into both national health plans and patient care pathways to drive a sustainable safety revolution in perioperative care.

View all citing articles on Scopus

View full text

The appropriateness of 30-day mortality as a quality metric in colorectal cancer surgery

Abstract

Background

Methods

Results

Conclusions

Introduction

Section snippets

Materials and methods

Results

Discussion

Conclusions

Disclosure

Funding

J Gastrointest Surg

Cancer statistics

CA Cancer J Clin

American Cancer Society Cancer Facts & Figures

Population-based cohort study comparing 30- and 90-day institutional mortality rates after colorectal surgery

Br J Surg

Hospital volume and surgical mortality in the United States

N Engl J Med

Doubling of 30-day mortality by 90 Days after esophagectomy: a critical measure of outcomes for quality improvement

Ann Surg

The 90-day mortality after pancreatectomy for cancer is double the 30-day mortality: more than 20,000 resections from the national cancer data base

Ann Surg Oncol