BACKGROUND: Randomized trials show that pneumatic dilation (PD) ≥30 mm and laparoscopic myotomy (LM) provide equivalent symptom relief and disease-related quality of life for patients with achalasia. However, questions remain about the safety, burden, and costs of treatment options. STUDY DESIGN: We performed a retrospective cohort study of achalasia patients initially treated with PD or LM (2009 to 2014) using the Truven Health MarketScan Research Databases. All patients had 1 year of follow-up after initial treatment. We compared safety, health care use, and total and out-of-pocket costs using generalized linear models. RESULTS: Among 1,061 patients, 82% were treated with LM. The LM patients were younger (median age 49 vs 52 years; p < 0.01), but were similar in terms of sex (p = 0.80) and prevalence of comorbid conditions (p = 0.11). There were no significant differences in the 1-year cumulative risk of esophageal perforation (LM 0.8% vs PD 1.6%; p = 0.32) or 30-day mortality (LM 0.3% vs PD 0.5%; p = 0.71). Laparoscopic myotomy was associated with an 82% lower rate of reintervention (p < 0.01), a 29% lower rate of subsequent diagnostic testing (p < 0.01), and a 53% lower rate of readmission (p < 0.01). Total and out-of-pocket costs were not significantly different (p > 0.05). CONCLUSIONS: In the US, LM appears to be the preferred treatment for achalasia. Both LM and PD appear to be safe interventions. Along a short time horizon, the costs of LM and PD were not different. Mirroring findings from randomized trials, LM is associated with fewer reinterventions, less diagnostic testing, and fewer hospitalizations.
5 February 2016 in Journal of the American College of Surgeons
Background Randomized trials have found that alvimopan hastens return of bowel function and reduces length of stay (LOS) by 1 day among patients undergoing colorectal surgery. However, its effectiveness in routine clinical practice and its impact on hospital costs remain uncertain. Study Design We performed a retrospective cohort study of patients undergoing elective colorectal surgery in Washington state (2009 to 2013) using data from a clinical registry (Surgical Care and Outcomes Assessment Program) linked to a statewide hospital discharge database (Comprehensive Hospital Abstract Reporting System). We used generalized estimating equations to evaluate the relationship between alvimopan and outcomes, and adjusted for patient, operative, and management characteristics. Hospital charges were converted to costs using hospital-specific charge to cost ratios, and were adjusted for inflation to 2013 US dollars. Results Among 14,781 patients undergoing elective colorectal surgery at 51 hospitals, 1,615 (11%) received alvimopan. Patients who received alvimopan had a LOS that was 1.8 days shorter (p < 0.01) and costs that were $2,017 lower (p < 0.01) compared with those who did not receive alvimopan. After adjustment, LOS was 0.9 days shorter (p < 0.01), and hospital costs were $636 lower (p = 0.02) among those receiving alvimopan compared with those who did not. Conclusions When used in routine clinical practice, alvimopan was associated with a shorter LOS and limited but significant hospital cost savings. Both efficacy and effectiveness data support the use of alvimopan in routine clinical practice, and its use could be measured as a marker of higher quality care.
20 April 2017 in eGEMs (Generating Evidence & Methods to improve patient outcomes)
Machine learning is used to analyze big data, often for the purposes of prediction. Analyzing a patient's healthcare utilization pattern may provide more precise estimates of risk for adverse events (AE) or death. We sought to characterize healthcare utilization prior to surgery using machine learning for the purposes of risk prediction. Patients from MarketScan Commercial Claims and Encounters Database undergoing elective surgery from 2007-2012 with ≥1 comorbidity were included. All available healthcare claims occurring within six months prior to surgery were assessed. More than 300 predictors were defined by considering all combinations of conditions, encounter types, and timing along with sociodemographic factors. We used a supervised Naive Bayes algorithm to predict risk of AE or death within 90 days of surgery. We compared the model's performance to the Charlson's comorbidity index, a commonly used risk prediction tool. Among 410,521 patients (mean age 52, 52 ± 9.4, 56% female), 4.7% had an AE and 0.01% died. The Charlson's comorbidity index predicted 57% of AE's and 59% of deaths. The Naive Bayes algorithm predicted 79% of AE's and 78% of deaths. Claims for cancer, kidney disease, and peripheral vascular disease were the primary drivers of AE or death following surgery. The use of machine learning algorithms improves upon one commonly used risk estimator. Precisely quantifying the risk of an AE following surgery may better inform patient-centered decision-making and direct targeted quality improvement interventions while supporting activities of accountable care organizations that rely on accurate estimates of population risk.
Background: Surgical site infection (SSI) after spine surgery is classified as a “never event” by the Centers for Medicare and Medicaid. Intra-wound antibiotics (IWA) have been proposed to reduce the incidence of SSI, but robust evidence to support its use is lacking. Methods: Prospective cohort undergoing spine fusion at 20 Washington State hospitals (July 2011 to March 2014) participating in the Spine Surgical Care and Outcomes Assessment Program (Spine SCOAP) linked to a discharge tracking system. Patient, hospital, and operative factors associated with SSI and IWA use during index hospitalizations through 600 days were analyzed using a random effects logistic model (index), and a time-to-event analysis (follow-up) using Cox proportional hazards. Results: A total of 9,823 patients underwent cervical (47%) or lumbar (53%) procedures (mean age, 58; 54% female) with an SSI rate of 1.1% during index hospitalization. Those with SSI were older, more often had diabetes mellitus, and more frequently underwent lumbar (versus cervical) fusion compared with those without SSI (all p < 0.01). Unadjusted rates of SSI during index hospitalization were lower in patients who received IWA (0.8% versus 1.5%). After adjustment for patient, hospital, and operative factors, no benefit was observed in those receiving IWA (odds ratio [OR] 0.65, 95% confidence interval [CI]: 0.42–1.03). At 12 mo, unadjusted rates of SSI were 2.4% and 3.0% for those who did and did not receive antibiotics; after adjustment there was no significant difference (hazard ratio [HR] 0.94, 95% CI: 0.62–1.42). Conclusions: Whereas unadjusted analyses indicate a nearly 50% reduction in index SSI using IWA, we did not observe a statistically significant difference after adjustment. Despite its size, this study is underpowered to detect small but potentially relevant improvements in rates of SSI. It remains to be determined if IWA should be promoted as a quality improvement intervention. Concerns related to bias in the use of IWA suggest the benefit of a randomized trial.
1 December 2016 in BMC Medical Informatics and Decision Making
Prospect theory suggests that when faced with an uncertain outcome, people display loss aversion by preferring to risk a greater loss rather than incurring certain, lesser cost. Providing probability information improves decision making towards the economically optimal choice in these situations. Clinicians frequently make decisions when the outcome is uncertain, and loss aversion may influence choices. This study explores the extent to which prospect theory, loss aversion, and probability information in a non-clinical domain explains clinical decision making under uncertainty. Four hundred sixty two participants (n = 117 non-medical undergraduates, n = 113 medical students, n = 117 resident trainees, and n = 115 medical/surgical faculty) completed a three-part online task. First, participants completed an iced-road salting task using temperature forecasts with or without explicit probability information. Second, participants chose between less or more risk-averse (“defensive medicine”) decisions in standardized scenarios. Last, participants chose between recommending therapy with certain outcomes or risking additional years gained or lost. In the road salting task, the mean expected value for decisions made by clinicians was better than for non-clinicians(−$1,022 vs -$1,061; <0.001). Probability information improved decision making for all participants, but non-clinicians improved more (mean improvement of $64 versus $33; p = 0.027). Mean defensive decisions decreased across training level (medical students 2.1 ± 0.9, residents 1.6 ± 0.8, faculty1.6 ± 1.1; p-trend < 0.001) and prospect-theory-concordant decisions increased (25.4%, 33.9%, and 40.7%;p-trend = 0.016). There was no relationship identified between road salting choices with defensive medicine and prospect-theory-concordant decisions. All participants made more economically-rational decisions when provided explicit probability information in a non-clinical domain. However, choices in the non-clinical domain were not related to prospect-theory concordant decision making and risk aversion tendencies in the clinical domain. Recognizing this discordance may be important when applying prospect theory to interventions aimed at improving clinical care.