The negative role of malnutrition in patients with Crohn’s disease is known; however, many coexisting disease-related factors could cause misinterpretation of the real culprit. This study aimed to describe the role of malnutrition using a novel methodology, entropy balancing. This was a retrospective analysis of consecutive patients undergoing elective major surgery for Crohn’s disease, preoperatively screened following the European Society for Clinical Nutrition guidelines. Two-step entropy balancing was applied to the group of malnourished patients to obtain an equal cohort having a null or low risk of malnutrition. The first reweighting homogenised the cohorts for non-modifiable confounding factors. The second reweighting matched the two groups for modifiable nutritional factors, assuming successful treatment of malnutrition. The entropy balancing was evaluated using the d-value. Postoperative results are reported as mean difference or OR, with a 95 % CI. Of the 183 patients, 69 (37·7 %) were at moderate/high risk for malnutrition. The malnourished patients had lower BMI (d = 1·000), Hb (d = 0·715), serum albumin (d = 0·981), a higher lymphocyte count (d = 0·124), Charlson Comorbidity Index (d = 0·257), American Society of Anaesthesiologists (d = 0·327) and Harvey-Bradshaw scores (d = 0·696). Protective loop ileostomy was more frequently performed (d = 0·648) in the malnourished group. After the first reweighting, malnourished patients experienced a prolonged length of stay (mean difference = 1·9; 0·11, 3·71, days), higher overall complication rate (OR 4·42; 1·39, 13·97) and higher comprehensive complication index score (mean difference = 8·9; 2·2 15·7). After the second reweighting, the postoperative course of the two groups was comparable. Entropy balancing showed the independent role of preoperative malnutrition and the possible advantages obtainable from a pre-habilitation programme in Crohn’s disease patients awaiting surgery.

We introduce hierarchically regularized entropy balancing as an extension to entropy balancing, a reweighting method that adjusts weights for control group units to achieve covariate balance in observational studies with binary treatments. Our proposed extension expands the feature space by including higher-order terms (such as squared and cubic terms and interactions) of covariates and then achieves approximate balance on the expanded features using ridge penalties with a hierarchical structure. Compared with entropy balancing, this extension relaxes model dependency and improves the robustness of causal estimates while avoiding optimization failure or highly concentrated weights. It prevents specification searches by minimizing user discretion in selecting features to balance on and is also computationally more efficient than kernel balancing, a kernel-based covariate balancing method. We demonstrate its performance through simulations and an empirical example. We develop an open-source R package, hbal, to facilitate implementation.

We compare the trajectories of families who have a child with poor neonatal health compared to those who do not, using administrative birth record data merged with longitudinal household survey data. We apply entropy balancing and weighting methods to enhance comparison between the two types of families. We find that children with poor neonatal health are more likely to be diagnosed with a disability and to receive Supplemental Security Income. Mothers who have children with poor neonatal health are more likely to reduce labor force participation at both the intensive and extensive margins when the child is young. Further, they are more likely to receive benefits from the Supplemental Nutrition Assistance Program and the Temporary Assistance for Needy Families program. We find no significant effects of poor neonatal health on maternal mental health or household income. Parents who have children with poor neonatal health are significantly less likely to remain married or cohabiting.