Statewide geographic variation in hysterectomy approach for pelvic organ prolapse: a county-level analysis.

BACKGROUND: There are no definitive guidelines for surgical treatment of pelvic organ prolapse. Previous data suggests geographic variation in apical repair rates in health systems throughout the United States. Such variation can reflect lack of standardized treatment pathways. An additional area of variation for pelvic organ prolapse repair may be hysterectomy approach which could not only influence concurrent repair procedures, but also healthcare utilization. OBJECTIVE: This study aimed to examine statewide geographic variation in surgical approach of hysterectomy for prolapse repair and concurrent use of colporrhaphy and colpopexy. STUDY DESIGN: We conducted a retrospective analysis of Blue Cross Blue Shield, Medicare, and Medicaid fee-for-service insurance claims for hysterectomies performed for prolapse in Michigan between October 2015 and December 2021. Prolapse was identified with International Classification of Disease Tenth Revision codes. The primary outcome was variation in surgical approach for hysterectomy as determined by Current Procedural Terminology code (vaginal, laparoscopic, laparoscopic assisted vaginal, or abdominal) on a county level. Patient home address zip codes were used to determine county of residence. A hierarchical multivariable logistic regression model with vaginal approach as the dependent variable and county-level random effects was estimated. Patient attributes, including age, comorbidities (diabetes mellitus, chronic obstructive pulmonary disease, congestive heart failure, morbid obesity), concurrent gynecologic diagnoses, health insurance type, and social vulnerability index were used as fixed-effects. To estimate variation between counties in vaginal hysterectomy rates, a median odds ratio was calculated. RESULTS: There were 6974 hysterectomies for prolapse representing 78 total counties that met eligibility criteria. Of these, 2865 (41.1%) underwent vaginal hysterectomy, 1119 (16.0%) underwent laparoscopic assisted vaginal hysterectomy, and 2990 (42.9%) underwent laparoscopic hysterectomy. The proportion of vaginal hysterectomy across 78 counties ranged from 5.8% to 86.8%. The median odds ratio was 1.86 (95% credible interval, 1.33-3.83), consistent with a high level of variation. Thirty-seven counties were considered statistical outliers because the observed proportion of vaginal hysterectomy was outside the predicted range (as defined by confidence intervals of the funnel plot). Vaginal hysterectomy was associated with higher rates of concurrent colporrhaphy than laparoscopic assisted vaginal hysterectomy or laparoscopic hysterectomy (88.5% vs 65.6% vs 41.1%, respectively; P<.001) and lower rates of concurrent colpopexy (45.7% vs 51.7% vs 80.1%, respectively; P<.001). CONCLUSION: This statewide analysis reveals a significant level of variation in the surgical approach for hysterectomies performed for prolapse. The variation in surgical approach for hysterectomy may help account for high rates of variation in concurrent procedures, especially apical suspension procedures. These data highlight how geographic location may influence the surgical procedures a patient undergoes for uterine prolapse.

Outcomes of Elective Induction of Labor at 39 Weeks from a Statewide Collaborative Quality Initiative.

OBJECTIVE: This article evaluates the impact of adopting a practice of elective induction of labor (eIOL) at 39 weeks among nulliparous, term, singleton, vertex (NTSV) pregnancies in a statewide collaborative. STUDY DESIGN: We used data from a statewide maternity hospital collaborative quality initiative to analyze pregnancies that reached 39 weeks without a medical indication for delivery. We compared patients who underwent an eIOL versus those who experienced expectant management. The eIOL cohort was subsequently compared with a propensity score-matched cohort who were expectantly managed. The primary outcome was cesarean birth rate. Secondary outcomes included time to delivery and maternal and neonatal morbidities. Chi-square test, t-test, logistic regression, and propensity score matching methods were used for analysis. RESULTS: In 2020, 27,313 NTSV pregnancies were entered into the collaborative's data registry. A total of 1,558 women underwent eIOL and 12,577 were expectantly managed. Women in the eIOL cohort were more likely to be ≥35 years old (12.1 vs. 5.3%, p < 0.001), identify as white non-Hispanic (73.9 vs. 66.8%, p < 0.001), and be privately insured (63.0 vs. 61.3%, p = 0.04). When compared with all expectantly managed women, eIOL was associated with a higher cesarean birth rate (30.1 vs. 23.6%, p < 0.001). When compared with a propensity score-matched cohort, eIOL was not associated with a difference in cesarean birth rate (30.1 vs. 30.7%, p = 0.697). Time from admission to delivery was longer for the eIOL cohort compared with the unmatched (24.7 ± 12.3 vs. 16.3 ± 11.3 hours, p < 0.001) and matched (24.7 ± 12.3 vs. 20.1 ± 12.0 hours, p < 0.001) cohorts. Expectantly managed women were less likely to have a postpartum hemorrhage (8.3 vs. 10.1%, p = 0.02) or operative delivery (9.3 vs. 11.4%, p = 0.029), whereas women who underwent an eIOL were less likely to have a hypertensive disorder of pregnancy (5.5 vs. 9.2%, p < 0.001). CONCLUSION: eIOL at 39 weeks may not be associated with a reduced NTSV cesarean delivery rate. KEY POINTS: · Elective IOL at 39 weeks may not be associated with a reduced NTSV cesarean delivery rate.. · The practice of elective induction of labor may not be equitably applied across birthing people.. · Further research is needed to identify best practices to support people undergoing labor induction..

Variation in Opioid Prescribing After Vaginal and Cesarean Birth: A Statewide Analysis.

INTRODUCTION: Our aim was to evaluate variation in opioid prescribing rates and prescription size following childbirth across providers and hospitals. METHODS: This retrospective cohort study analyzed claims data from a single-payer Preferred Provider Organization from June 2014 to May 2019 in 84 hospitals in a statewide quality collaborative. All patients aged 12-55 years, undergoing childbirth, with continuous enrollment in pregnancy were included. The primary outcome was the predicted rate of postpartum opioid fills from 7 days before birth to 3 days after discharge. Secondary outcomes included postpartum opioid prescription size in oral morphine equivalents, a standardized measure that includes the number of pills prescribed times the strength of the medication. Multilevel regression models accounted for clustering. We calculated attributable variation in opioid fills using the intraclass correlation coefficient. RESULTS: Of 41,427 births, 15,459 patients (37.2%) filled a postpartum opioid prescription (vaginal, 4,624/27,536 [16.8%]; cesarean, 10,835/13,891 [78.0%]). The median postpartum prescription size was 150 oral morphine equivalents (interquartile range [IQR], 30) (vaginal, 135; [IQR, 45]; cesarean, 150 [IQR, 75]). In adjusted models, the rates of opioid prescribing after vaginal birth differed from cesarean birth (vaginal median, 12.1% [range, 1.1%-60.0%]; cesarean median, 80.4% [range, 43.6%-90.2%]). More variation in postpartum opioid fills was attributable to providers and hospitals for vaginal (provider, 29%; hospital, 24%) than cesarean birth (provider, 8%; hospital, 6%). Variation in prescription size was driven by providers for vaginal birth (provider, 27%; hospital, 6%) and providers and hospitals for cesarean birth (provider, 29%; hospital, 21%). CONCLUSIONS: Across a statewide quality collaborative, variation in postpartum opioid prescribing is attributable to providers and hospitals. Future efforts at the provider and hospital levels are needed to implement best practices for postpartum opioid prescribing.

Predictors of same-day discharge following benign minimally invasive hysterectomy.

BACKGROUND: Same-day discharge following minimally invasive hysterectomy has been shown to be safe and feasible in select populations, but many nonclinical factors influencing same-day discharge remain unexplored. OBJECTIVE: To develop prediction models for same-day discharge following minimally invasive hysterectomy using both clinical and nonclinical attributes and to compare model concordance of individual attribute groups. STUDY DESIGN: We performed a retrospective study of patients who underwent elective minimally invasive hysterectomy for benign gynecologic indications at 69 hospitals in a statewide quality improvement collaborative between 2012 and 2019. Potential predictors of same-day discharge were determined a priori and placed into 1 of 7 attribute groupings: intraoperative, surgeon, hospital, surgical timing, patient clinical, patient socioeconomic, and patient geographic attributes. To account for clustering of same-day discharge practices among surgeons and within hospitals, hierarchical multivariable logistic regression models were fitted using predictors from each attribute group individually and all predictors in a composite model. Receiver operator characteristic curves were generated for each model. The Hanley-McNeil test was used for comparisons, 95% confidence intervals for the areas under the receiver operator characteristic curve were calculated, and a P value of <.05 was considered significant. RESULTS: Of the 23,513 patients in our study, 5062 (21.5%) had same-day discharge. The composite model had an area under the receiver operator characteristic curve of 0.770 (95% confidence interval, 0.763-0.777). Among models using factors from individual attribute groups, the model using intraoperative attributes had the highest concordance for same-day discharge (area under the receiver operator characteristic curve, 0.720; 95% confidence interval, 0.712-0.727). The models using surgeon and hospital attributes were the second and third most concordant, respectively (area under the receiver operator characteristic curve, 0.678; 95% confidence interval, 0.670-0.685; area under the receiver operator characteristic curve, 0.655; 95% confidence interval, 0.656-0.664). Models using surgical timing and patient clinical, socioeconomic, and geographic attributes had poor predictive ability (all areas under the receiver operator characteristic curve <0.6). CONCLUSION: Clinical and nonclinical attributes contributed to a composite prediction model with good discrimination in predicting same-day discharge following minimally invasive hysterectomy. Factors related to intraoperative, hospital, and surgeon attributes produced models with the strongest predictive ability. Focusing on these attributes may aid efforts to improve utilization of same-day discharge following minimally invasive hysterectomy.

Effects of Pharmacologic Venous Thromboembolism Prophylaxis in Benign Hysterectomy.

STUDY OBJECTIVE: To evaluate whether the addition of pharmacologic prophylaxis to mechanical prophylaxis for venous thromboembolism (VTE) is associated with changes in perioperative outcomes in hysterectomy for benign indications. DESIGN: Retrospective cohort study. SETTING: Michigan Surgical Quality Collaborative database. PATIENTS: Patients who underwent hysterectomy between July 2012 and June 2015 when VTE prophylaxis data were collected. INTERVENTIONS: Patients who received mechanical prophylaxis alone were compared with those receiving dual prophylaxis (mechanical and pharmacologic). Minimally invasive surgeries (MIS) included laparoscopic, vaginal, robotic-assisted, and laparoscopic-assisted vaginal hysterectomies and were analyzed separately from abdominal (ABD) hysterectomy. MEASUREMENTS AND MAIN RESULTS: Propensity score matching was used to minimize confounding because of the differences in demographic and perioperative characteristics. The primary outcome was estimated blood loss (EBL). The secondary outcomes were operative time, postoperative blood transfusion, VTE, surgical site infection, reoperation, readmission, and death. There were 1803 matched pairs in the MIS analysis. In the ABD hysterectomy analysis, 2:1 matching was used with a total of 1168 patients receiving mechanical prophylaxis alone matched to 616 patients receiving dual prophylaxis. EBL was higher by 54.5 mL (95% confidence interval [CI], 16.9-92.1) in those receiving dual prophylaxis in the ABD hysterectomy analysis but did not differ between groups in the MIS analysis. Operative time was significantly longer with dual prophylaxis in both MIS (18.3 minutes; 95% CI, 13.8-22.8) and ABD (15.3 minutes; 95% CI, 9.0-21.6) surgical approaches. There was no difference in other secondary outcomes. CONCLUSION: The addition of pharmacologic prophylaxis to mechanical prophylaxis in benign hysterectomy was associated with longer operative time, regardless of surgical approach and increased EBL in ABD hysterectomy. Given very low rates of VTE, no difference in other perioperative outcomes, and possible harm, it seems reasonable to encourage individualized rather than routine use of pharmacologic prophylaxis in patients undergoing benign hysterectomy receiving mechanical prophylaxis.

Development of a Preoperative Clinical Risk Assessment Tool for Postoperative Complications After Hysterectomy.

STUDY OBJECTIVE: To develop a preoperative risk assessment tool that quantifies the risk of postoperative complications within 30 days of hysterectomy. DESIGN: Retrospective analysis. SETTING: Michigan Surgical Quality Collaborative hospitals. PATIENTS: Women who underwent hysterectomy for gynecologic indications. INTERVENTIONS: Development of a nomogram to create a clinical risk assessment tool. MEASUREMENTS AND MAIN RESULTS: Postoperative complications within 30 days were the primary outcome. Bivariate analysis was performed comparing women who had a complication and those who did not. The patient registry was randomly divided. A logistic regression model developed and validated from the Collaborative database was externally validated with hysterectomy cases from the National Surgical Quality Improvement Program, and a nomogram was developed to create a clinical risk assessment tool. Of the 41,147 included women, the overall postoperative complication rate was 3.98% (n = 1638). Preoperative factors associated with postoperative complications were sepsis (odds ratio [OR] 7.98; confidence interval [CI], 1.98-32.20), abdominal approach (OR 2.27; 95% CI, 1.70-3.05), dependent functional status (OR 2.20; 95% CI, 1.34-3.62), bleeding disorder (OR 2.10; 95% CI, 1.37-3.21), diabetes with HbA1c ≥9% (OR 1.93; 95% CI, 1.16-3.24), gynecologic cancer (OR 1.86; 95% CI, 1.49-2.31), blood transfusion (OR 1.84; 95% CI, 1.15-2.96), American Society of Anesthesiologists Physical Status Classification System class ≥3 (OR 1.46; 95% CI, 1.24-1.73), government insurance (OR 1.3; 95% CI, 1.40-1.90), and body mass index ≥40 (OR 1.25; 95% CI, 1.04-1.50). Model discrimination was consistent in the derivation, internal validation, and external validation cohorts (C-statistics 0.68, 0.69, 0.68, respectively). CONCLUSION: We validated a preoperative clinical risk assessment tool to predict postoperative complications within 30 days of hysterectomy. Modifiable risk factors identified were preoperative blood transfusion, poor glycemic control, and open abdominal surgery.

Vaginal preparation with povidone-iodine or chlorhexidine before hysterectomy: a propensity score matched analysis.

BACKGROUND: Antiseptic vaginal preparation is recommended before gynecologic surgery; however, there is a lack of data regarding the effectiveness of different agents. OBJECTIVE: To compare rates of postoperative infectious complications and hospital utilization with preoperative vaginal preparation using povidone-iodine or chlorhexidine before hysterectomy. STUDY DESIGN: This was a retrospective analysis of patients who underwent hysterectomy for gynecologic indications at 70 hospitals in a statewide surgical collaborative between January 2017 and December 2019. The primary outcome was postoperative infectious complications (including urinary tract infection, surgical site infections involving superficial, deep, or organ space tissues, or cellulitis) within 30 days of surgery. To adjust for confounding, propensity score matching, 1:1 without replacement and with a caliper of.005 was performed to create cohorts that had vaginal preparation with either povidone-iodine or chlorhexidine and did not differ in observable characteristics. We compared the rates of infectious morbidity and hospital utilization (emergency department visits, readmission, reoperation) in the matched cohorts. RESULTS: In the statewide collaborative, there were 18,184 patients who received povidone-iodine and 3018 who received chlorhexidine. After propensity score matching of 2935 pairs, the povidone-iodine and chlorhexidine groups did not differ in demographics, comorbidities, choice of preoperative antibiotics, benign vs malignant surgical indication, and surgical approach. Povidone-iodine was associated with a lower rate of infectious morbidity (3.0% vs 4.3%; P=.01), urinary tract infection (1.1% vs 1.7%; P=.03) and emergency department visits (7.9% vs 9.7%; P=.01) than with chlorhexidine. There were nonsignificant trends of lower rates of surgical site infection (2.0% vs 2.7%; P=.07) and reoperation (1.6% vs 2.1%; P=.18). CONCLUSION: This propensity score matched analysis provides evidence that povidone-iodine is preferable to chlorhexidine for vaginal preparation before hysterectomy because of lower rates of infectious morbidity and fewer emergency department visits. However, the absolute differences in infectious morbidity rates were approximately 1%, and in the event of an iodine allergy, chlorhexidine appears to be a reasonable alternative.

Hysterectomy Complications Relative to HbA1c Levels: Identifying a Threshold for Surgical Planning.

STUDY OBJECTIVE: To evaluate whether diabetes diagnosis and level of diabetes control as reflected by higher preoperative glycosylated hemoglobin (HbA1c) levels are associated with increased complication rates after hysterectomy and to identify a threshold of preoperative HbA1c level past which we should consider delaying surgery owing to increased risk of complications. DESIGN: Retrospective cohort study. SETTING: Hospitals in the Michigan Surgical Quality Collaborative between June 4, 2012, and October 17, 2017. PATIENTS: Women with and without a diabetes diagnosis. INTERVENTIONS: Hysterectomy. MEASUREMENTS AND MAIN RESULTS: Data on demographics, preoperative HbA1c values, surgical approach, composite postoperative complications, readmissions, emergency department visits, and reoperations were abstracted. The risk of a postoperative complication when diabetes was stratified by preoperative HbA1c level was evaluated in a sensitivity analysis, and independent associations were identified in a mixed, multivariate logistic regression model. We identified 41 286 hysterectomies performed at 70 hospitals to be included for analysis. The sensitivity analysis identified 4 groups of risk for postoperative complications: group 1: no diabetes diagnosis and no HbA1c value; group 2: no diabetes diagnosis, with HbA1c levels between 4% and 6.5%; group 3: diabetes diagnosis and no HbA1c value or HbA1c levels <9%; and group 4: diabetes diagnosis with HbA1c levels ≥9%. In the adjusted model, there were significant 32% and 34% increased odds of postoperative complications for groups 2 and 3, respectively, compared with group 1. There were more than 2-fold increased odds of complications for women with diabetes and a preoperative HbA1c level ≥9% (group 4) compared with the women in group 1. Diabetes diagnosis with preoperative HbA1c levels ≥9% had increased odds of complications compared with diabetes diagnosis with preoperative HbA1c levels <9%. Patients with well-controlled diabetes seemed to have increased odds of complications with laparoscopic surgery. CONCLUSION: Diabetes diagnosis and measurement of preoperative HbA1c levels provide risk stratification for postoperative complications after hysterectomy, with the highest observed effect among patients with diabetes with a preoperative HbA1c level ≥9%.

Impact of a National Private Health Insurer's Prior Authorization Policy on Utilization of Vaginal Hysterectomy.

OBJECTIVES: On April 6, 2015, the largest private health insurer in the United States implemented a policy requiring prior authorization for all hysterectomies except those done as outpatient vaginal. The purpose of this policy was to increase utilization of vaginal hysterectomy; however, it is unknown whether this policy had its intended effect. We sought to analyze trends in hysterectomy routes before and after implementation of the prior authorization policy to see if utilization of vaginal hysterectomy increased. METHODS: This was a retrospective study using the Optum Clinformatics Data Mart national claims database of women enrolled in a single national private health insurer who underwent hysterectomy for any indication between January 1, 2010, and June 30, 2016. Per-quarter utilization of hysterectomy routes (abdominal, laparoscopic, vaginal, and laparoscopic-assisted vaginal) was compared between the prepolicy and postpolicy periods using interrupted time series analyses. RESULTS: Data for 305,139 hysterectomies were available-248,821 in the prepolicy period and 56,318 in the postperiod. Outpatient vaginal hysterectomy had the greatest increase in utilization of all routes and types; the average utilization per quarter in the prepolicy period was -0.61%, and this increased to 0.21% in the postpolicy period (P < 0.0001). Outpatient laparoscopic hysterectomy had the greatest decrease in utilization, with an average decrease of -1.50% per quarter. CONCLUSIONS: The prior authorization policy was associated with a short-term increase in utilization of vaginal hysterectomy.

New persistent opioid use after acute opioid prescribing in pregnancy: a nationwide analysis.

OBJECTIVE: To evaluate the association between opioid prescribing during pregnancy and new persistent opioid use in the year following delivery. MATERIALS AND METHODS: This nationwide retrospective cohort study included patients aged 12-55 years in Optum's deidentified Clinformatics Data Mart Database who were undergoing vaginal delivery or cesarean delivery from 2008 to 2016, with continuous enrollment from 2 years before birth to 1 year postdischarge. Women were included if they were opioid naive in pregnancy (ie, did not fill an opioid prescription 2 years to 9 months before delivery) and did not undergo a procedure within the year after discharge. The exposure was filling an opioid prescription in pregnancy. The primary outcome was new persistent opioid use, defined as a pharmacy claim for ≥1 opioid prescription between 4 and 90 days postdischarge and ≥1 prescription between 91 and 365 days postdischarge. Clinical and demographic covariates were included. Analyses included descriptive statistics and multivariable logistic regression, adjusting for clinical and demographic covariates. RESULTS: Of 158,425 childbirths identified, 101,013 (63.8%) were by vaginal delivery and 57,412 (36.2%) cesarean delivery. Among all patients, 6.0% (9429) filled an opioid prescription during pregnancy. The factors associated with filling an opioid in pregnancy were having a nondelivery procedure in pregnancy (adjusted odds ratio, 9.60; 95% confidence interval, 8.81-10.47) and having an emergency room visit during pregnancy (adjusted odds ratio, 2.48; 95% confidence interval, 2.37-2.59). Of women who received an opioid in pregnancy, 4% (379) developed new persistent opioid use. The factors most associated with new persistent opioid use were receiving an opioid prescription during pregnancy (adjusted odds ratio, 3.45; 95% confidence interval, 3.04-3.92) and filling a peripartum opioid prescription (1 week prior to 3 days postdischarge) adjusted odds ratio, 2.28, 95% confidence interval (2.02-2.57). Though having a procedure during pregnancy was associated with increased receipt of an opioid prescription, it was also associated with reduced new persistent opioid use (adjusted odds ratio, 0.72; 95% confidence interval, 0.52-0.99). CONCLUSION: Women who receive an opioid prescription during pregnancy are more likely to experience new persistent opioid use. Maternity care providers must balance pain management in pregnancy with potential risks of opioids.

Safety Bundles in Gynecology.

Patient safety bundles and checklists have been shown to improve outcomes in medicine, surgery, and obstetrics. Until recently, there has been less study into their use in the gynecology setting. Here, we review the available evidence and examples of successful checklist and bundle implementation in gynecology and encourage more robust implementation and standardization in our field going forward.

Better late than never: why obstetricians must implement enhanced recovery after cesarean.

Despite persistent concerns about high cesarean delivery rates internationally, there has been less attention on improving perioperative outcomes for the millions of women who will experience a cesarean delivery each year. Enhanced recovery after surgery, a standardized, evidence-based, interdisciplinary protocol, has been successfully used in other surgical specialties including gynecology to improve quality of care and patient satisfaction while reducing overall health care costs through reduced length of stay. Enhanced recovery after surgery society guidelines for cesarean delivery were just released in August 2018. Obstetric patients, who face the dual challenge of being postpartum and postoperative, could benefit greatly from protocols that optimize their return to physiological function and reduce surgical morbidity. Although enhanced recovery after surgery has been widespread in other surgical specialties, uptake of this protocol in obstetrics has lagged behind. We believe enhanced recovery after surgery for cesarean delivery can effectively address 3 challenges faced by obstetrician/gynecologists. These are: (1) improving care for the high number of women undergoing cesarean deliveries; (2) using evidence-based care bundles to prevent maternal morbidity and mortality, address disparities, and reduce costs; and (3) limiting postoperative opioid prescribing in response to the opioid crisis. Enhanced recovery after surgery for cesarean delivery and other standardized care protocols have the potential to reduce the disproportionately high rates of maternal morbidity and mortality in the United States, and ensure all patients, regardless of demographics or location, receive the same level of high-quality peripartum care.

Hospital contribution to variation in rates of vaginal birth after cesarean.

OBJECTIVE: To determine the influence of delivery hospital on the rate of vaginal birth after cesarean (VBAC). STUDY DESIGN: This retrospective cohort study used claims data from Blue Cross and Blue Shield of Michigan. Women with a prior cesarean and a singleton livebirth between 2012 and 2016 were included. We calculated the hospital-specific risk-standardized VBAC rates and median odds ratio as a measure of variation. RESULT: Hospital-level adjusted rates varied nearly tenfold (3.7%-35.5%). Compared to the lowest volume hospitals (1st quartile), the likelihood of VBAC increased for those in the 2nd (adjusted OR 2.75 [95% CI 1.23-6.17]), 3rd (adjusted OR 3.73 [95% CI 1.59-8.75]), and 4th quartiles (adjusted OR 2.9 [95% CI 1.11-7.72]). The median OR suggested significant variation by hospital after adjustment. CONCLUSION: The delivery hospital itself explains a large amount of the variation in rates of VBAC after adjustment for patient and hospital characteristics.

Importance of Estimated Blood Loss in Resource Utilization and Complications of Hysterectomy for Benign Indications.

OBJECTIVE: To identify the variation in estimated blood loss at the time of hysterectomy for benign indications and to analyze how blood loss is associated with measures of resource utilization and complications. METHODS: We conducted a retrospective cohort study and analyzed hysterectomy for benign indications at hospitals in the Michigan Surgical Quality Collaborative between January 1, 2013, and May 30, 2015. A sensitivity analysis was performed to identify how estimated blood loss was associated with measures of utilization (transfusion, readmission, reoperation, and length of stay) and postoperative complications. A hierarchical logistic regression model was used to identify patient level factors independently associated with estimated blood loss greater than 400 mL and to calculate a risk- and reliability-adjusted rate for each hospital. RESULTS: There were 18,033 hysterectomies for benign indications from 61 hospitals included for analysis. The median estimated blood loss was 100 mL, and the 90th percentile estimated blood loss was 400 mL. A sensitivity analysis demonstrated increased risks of transfusion, readmission, reoperation, length of stay, and major postoperative complications with estimated blood loss greater than 400 mL. The proportion of hysterectomies at hospitals in the collaborative with estimated blood loss greater than 400 mL ranged from 3.5% to 16.9% after risk and reliability adjustments. The risk factors with the highest adjusted odds for estimated blood loss greater than 400 mL included abdominal surgery compared with laparoscopic hysterectomy (adjusted odds ratio [aOR] 2.8, CI 2.3-3.5), surgical time longer than 3 hours (aOR 3.9, CI 3.3-4.5), and specimen weight greater than 250 g compared with less than 100 g (aOR 4.8, CI 3.9-5.8). Adhesive disease, low surgeon volume, being younger than 40 years of age, having a body mass index greater than 35, and the need for a preoperative transfusion were also statistically significantly associated with estimated blood loss greater than 400 mL. CONCLUSION: There is fivefold variation in the hospital rate of hysterectomies with an estimated blood loss greater than 400 mL (90th percentile)-a threshold associated with significantly higher rates of health care utilization and complications. Avoidance of abdominal hysterectomy when possible may reduce intraoperative blood loss and associated sequelae.

Documenting pessary offer prior to hysterectomy for management of pelvic organ prolapse.

OBJECTIVES: To (1) determine the proportion of hysterectomy cases with documentation of pessary counseling prior to prolapse surgery and (2) identify variables associated with women offered a pessary. STUDY DESIGN: The Michigan Surgical Quality Collaborative (MSQC) is a hysterectomy improvement initiative. Hysterectomies from 2013 to 2015 in which prolapse was the principal diagnosis were included. "Pessary offer" was defined as documentation showing the patient declined, could not tolerate, or failed a pessary trial. Bivariate analyses were used to compare demographics, medical history, surgical route, concomitant procedures (colpopexy or colporrhaphy), and intra- and postoperative complications between women with and without pessary offer. Hierarchical logistic regression was used to determine factors independently associated with pessary offer. Risk-adjusted pessary offer rates by hospital were calculated. RESULTS: The adjusted rate of pessary offer was 25.2%, ranging from 3 to 76% per hospital. Bivariate comparisons showed differences between women with and without pessary offer in age, tobacco use, prior pelvic surgery, insurance status, surgical approach, secondary indication for surgery, concomitant prolapse procedure, teaching hospital status and hospital bed size. In logistic regression, odds of pessary offer increased with age > 55 years (OR 1.45, 95% CI 1.12-1.88, p = 0.006), Medicare insurance (OR 1.65, 95% CI 1.30-2.10, p < 0.0001), and a concomitant procedure (OR 1.5, 95% CI 1.16-1.93, p = 0.002). Postoperative urinary tract infections were more common in patients offered a pessary (6.4% vs. 2.5%, p < 0.0001), but other complications were similar. CONCLUSIONS: Overall, only one-quarter of hysterectomies for prolapse in MSQC hospitals had documentation of pessary counseling-suggesting an opportunity to improve documentation, counseling regarding pessary use, or both.

Preoperatively predicting non-home discharge after surgery for gynecologic malignancy.

OBJECTIVE: Returning home after surgery is a desirable patient-centered outcome associated with decreased costs compared to non-home discharge. Our objective was to develop a preoperative risk-scoring model predicting non-home discharge after surgery for gynecologic malignancy. METHODS: Women who underwent surgery involving hysterectomy for gynecologic malignancy from 2013 to 2015 were identified from the Michigan Surgical Quality Collaborative database. Patients were divided by discharge destination, and a multivariable logistic regression model was developed to create a nomogram to assign case-specific risk scores. The model was validated using the National Surgical Quality Improvement Program (NSQIP) database. RESULTS: Non-home discharge occurred in 3.1% of 2134 women. The proportion of non-home discharges did not differ by cancer diagnosis (uterine 3.5%, ovarian 2.5%, and cervical 1.6%, p = 0.2). Skilled nursing facilities were the most common non-home destination (68.2%). Among patients with comorbidities (hypertension, diabetes, coronary artery disease, chronic obstructive pulmonary disease /dyspnea, arrhythmia, and history of deep vein thrombosis/pulmonary embolism), non-home discharge was more common in women with 1 (adjusted OR [aOR] 3.4; p = 0.03) or ≥2 of these comorbidities (aOR 5.1; p = 0.003) compared to none. Non-home discharge was more common after laparotomy (aOR 6.7; p < 0.0001) than laparoscopy, and in those aged ≥70 years (aOR 3.4; p < 0.0001) with American Society of Anesthesiologists class ≥ 3 (aOR 4.5; p = 0.0004) and dependent functional status (aOR 8.7; p < 0.0001). The model C-statistic was 0.89. When the model was applied to 4248 eligible patients from the NSQIP dataset, the C-statistic was 0.84 (95% CI: 0.79-0.89). CONCLUSIONS: Non-home discharge after surgery for gynecologic malignancy was predicted with high accuracy in this retrospective analysis.

Evaluation of the Methods Used by Medicare's Hospital-Acquired Condition Reduction Program to Identify Outlier Hospitals for Surgical Site Infection.

BACKGROUND: The Hospital Acquired Condition Reduction Program (HACRP) is a national pay-for-performance program that includes a measure of surgical site infection (SSI) after hysterectomy and colectomy. This study compares the HACRP SSI measure with other published methods. STUDY DESIGN: This was a retrospective cohort study from the Michigan Surgical Quality Collaborative (MSQC). The outcome was 30-day, adjusted deep and organ space SSI ("complex SSI"). Observed-to-expected ratios of complex SSI for each hospital were calculated using HACRP, National Healthcare Safety Network (NHSN), and MSQC methodologies. C-statistics were compared between models. Hospital rankings were compared, and ladder plots show changes in hospitals' HACRP scores that derive from each algorithm. RESULTS: Complex SSI occurred in 1.1% (190 of 16,672) of hysterectomies and 4.8% (n = 514 of 10,725) of colectomies. The HACRP risk-adjustment model for hysterectomy had a C-statistic of 0.55, significantly lower than NHSN (0.61, p = 0.0461) or MSQC models (0.77, p < 0.0001). For colectomy, C-statistics were 0.57, 0.66 (p < 0.0001) and 0.73 (p < 0.0001), respectively. For both operations, there were 5 high-outlier hospitals using HACRP, but fewer (4 or 3) using the other methods. Most hospitals in the bottom quartile were not statistical outliers, but would be flagged under HACRP. More than 50% of hospitals changed ranking position between models, which would result in different scores under HACRP. CONCLUSIONS: This study showed that the HACRP SSI measure unfairly places hospitals at risk for financial penalties that are not statistical outliers. Policy makers need to weigh the burden of data collection and the accuracy needed to identify hospitals for financial reward or penalty.