Concept: Propensity score matching
When estimating the average effect of a binary treatment (or exposure) on an outcome, methods that incorporate propensity scores, the G-formula, or targeted maximum likelihood estimation (TMLE) are preferred over naïve regression approaches, which are biased under misspecification of a parametric outcome model. In contrast propensity score methods require the correct specification of an exposure model. Double-robust methods only require correct specification of either the outcome or the exposure model. Targeted maximum likelihood estimation is a semiparametric double-robust method that improves the chances of correct model specification by allowing for flexible estimation using (nonparametric) machine-learning methods. It therefore requires weaker assumptions than its competitors. We provide a step-by-step guided implementation of TMLE and illustrate it in a realistic scenario based on cancer epidemiology where assumptions about correct model specification and positivity (ie, when a study participant had 0 probability of receiving the treatment) are nearly violated. This article provides a concise and reproducible educational introduction to TMLE for a binary outcome and exposure. The reader should gain sufficient understanding of TMLE from this introductory tutorial to be able to apply the method in practice. Extensive R-code is provided in easy-to-read boxes throughout the article for replicability. Stata users will find a testing implementation of TMLE and additional material in the Appendix S1 and at the following GitHub repository: https://github.com/migariane/SIM-TMLE-tutorial.
OBJECTIVE:: In a large nationwide administrative database of hospitalized patients, we investigated postoperative outcomes after laparoscopic or open distal gastrectomy in Japan. BACKGROUND:: The benefits of laparoscopic gastrectomy, such as decreased length of stay and morbidity, have typically been evaluated only with limited data on the basis of small samples. METHODS:: Using the Japanese Diagnosis Procedure Combination Database, we identified 9388 patients who were preoperatively diagnosed with stage I and II gastric cancer and underwent laparoscopic (n = 3937) or open (n = 5451) distal gastrectomy between July and December 2010. One-to-one propensity score matching was performed to compare in-hospital mortality, postoperative complication rates, length of stay, total costs, and 30-day readmission rates between the 2 groups. RESULTS:: Patients with younger age, lower comorbidity index, or stage I cancer were more likely to receive laparoscopic gastrectomy. In the propensity-matched analysis with 2473 pairs, the laparoscopic gastrectomy group in comparison with the open gastrectomy group showed a slight reduction in median postoperative length of stay (13 days vs 15 days, P < 0.001) but a slight increase in median total costs (US $21,510 vs $21,024, P = 0.002). There were no significant differences in in-hospital mortality (0.36% vs 0.28%, P = 0.80), overall postoperative complications (12.9% vs 12.6%, P = 0.73), or 30-day readmission rates (3.2% vs 3.2%, P = 0.94). CONCLUSIONS:: In this large nationwide cohort of patients with early-stage gastric cancer, laparoscopic gastrectomy was associated with a statistically significant but slight reduction in postoperative length of stay, but no differences between laparoscopic gastrectomy and open gastrectomy were detected in terms of early mortality and morbidity.
The introduction of complete mesocolic excision (CME) with central vessel ligation (CVL) for right-sided colon cancer has improved oncologic outcomes. However, there is controversy over the oncologic safety of laparoscopic CME with CVL. This study compared short-term and long-term oncologic outcomes between laparoscopic and open modified CME (mCME) with CVL in patients with right-sided colon cancer.
Limited treatment options are available for patients with hepatocellular carcinoma (HCC) with portal vein thrombosis (PVT). Transarterial radioembolization using Yttrium-90 microspheres is a new treatment modality for HCC with PVT. For this analysis, we compared responses to treatment with radioembolization and with sorafenib.
In England, clinicians and professional organisations report that higher numbers of adolescents with more severe psychosocial difficulties are accessing specialist services. A lack of national data on patterns of access to specialist services means there is limited information to inform policy. We examined whether severity of psychosocial difficulties in adolescents accessing mental healthcare has changed over time. Adolescents seen in specialist child mental healthcare in 2009 vs. 2014 were matched on demographics and problem types using propensity score matching; final sample N = 2776 adolescents. We found: 1) stability over time in overall severity of difficulties, 2) an increase in severity of young women’s emotional problems, and 3) a decrease in adolescents' conduct problems. The findings suggest the intriguing possibility that the criteria for accessing mental healthcare are not universally rising, but rather the patterns in access to specialist services may mirror epidemiological changes in severity of psychosocial difficulties in the population.
The care received by people presenting to hospital following self-harm varies and it is unclear how different types of treatment affect risk of further self-harm.
Double-adjustment can be used to remove confounding if imbalance exists after propensity score (PS) matching. However, it is not always possible to include all covariates in adjustment. We aimed to find the optimal imbalance threshold for entering covariates into regression.
Propensity scores (PS) are an increasingly popular method to adjust for confounding in observational studies. Propensity score methods have theoretical advantages over conventional covariate adjustment, but their relative performance in real-word scenarios is poorly characterized. We used datasets from 4 large-scale cardiovascular observational studies (PROMETHEUS, ADAPT-DES [the Assessment of Dual AntiPlatelet Therapy with Drug-Eluting Stents], THIN [The Health Improvement Network], and CHARM [Candesartan in Heart Failure-Assessment of Reduction in Mortality and Morbidity]) to compare the performance of conventional covariate adjustment with 4 common PS methods: matching, stratification, inverse probability weighting, and use of PS as a covariate. We found that stratification performed poorly with few outcome events, and inverse probability weighting gave imprecise estimates of treatment effect and undue influence to a small number of observations when substantial confounding was present. Covariate adjustment and matching performed well in all of our examples, although matching tended to give less precise estimates in some cases. PS methods are not necessarily superior to conventional covariate adjustment, and care should be taken to select the most suitable method.
The Tokyo guidelines recommend initial cholecystostomy tube drainage, antibiotics, and delayed cholecystectomy in patients with grade III cholecystitis.
Many studies have found a statistical association between breastfeeding and childhood adiposity. This paper investigates whether breastfeeding has an effect on subsequent childhood body mass index (BMI) using propensity scores to account for confounding.