Investigating Bias from Missing Data in an Electronic Health Records-Based Study of Weight Loss After Bariatric Surgery

Factors Associated with Missing Sociodemographic Data in the IRIS® (Intelligent Research in Sight) Registry

Purpose

To describe the prevalence of missing sociodemographic data in the IRIS® (Intelligent Research in Sight) Registry and to identify practice-level characteristics associated with missing sociodemographic data.

Design

Cross-sectional study.

Participants

All patients with clinical encounters at practices participating in the IRIS Registry prior to December 31, 2020.

Methods

We describe geographic and temporal trends in the prevalence of missing data for each sociodemographic variable (age, sex, race, ethnicity, geographic location, insurance type, and smoking status). Each practice contributing data to the registry was categorized based on the number of patients, number of physicians, geographic location, patient visit frequency, and patient population demographics.

Main Outcome Measures

Multivariable linear regression was used to describe the association of practice-level characteristics with missing patient-level sociodemographic data.

Results

This study included the electronic health records of 66 477 365 patients receiving care at 3306 practices participating in the IRIS Registry. The median number of patients per practice was 11 415 (interquartile range: 5849-24 148) and the median number of physicians per practice was 3 (interquartile range: 1-7). The prevalence of missing patient sociodemographic data were 0.1% for birth year, 0.4% for sex, 24.8% for race, 30.2% for ethnicity, 2.3% for 3-digit zip code, 14.8% for state, 5.5% for smoking status, and 17.0% for insurance type. The prevalence of missing data increased over time and varied at the state-level. Missing race data were associated with practices that had fewer visits per patient (P < 0.001), cared for a larger nonprivately insured patient population (P = 0.001), and were located in urban areas (P < 0.001). Frequent patient visits were associated with a lower prevalence of missing race (P < 0.001), ethnicity (P < 0.001), and insurance (P < 0.001), but a higher prevalence of missing smoking status (P < 0.001).

Conclusions

There are geographic and temporal trends in missing race, ethnicity, and insurance type data in the IRIS Registry. Several practice-level characteristics, including practice size, geographic location, and patient population, are associated with missing sociodemographic data. While the prevalence and patterns of missing data may change in future versions of the IRIS registry, there will remain a need to develop standardized approaches for minimizing potential sources of bias and ensure reproducibility across research studies.

Financial Disclosures

Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.

Missing data matter: an empirical evaluation of the impacts of missing EHR data in comparative effectiveness research

OBJECTIVES

The impacts of missing data in comparative effectiveness research (CER) using electronic health records (EHRs) may vary depending on the type and pattern of missing data. In this study, we aimed to quantify these impacts and compare the performance of different imputation methods.

MATERIALS AND METHODS

We conducted an empirical (simulation) study to quantify the bias and power loss in estimating treatment effects in CER using EHR data. We considered various missing scenarios and used the propensity scores to control for confounding. We compared the performance of the multiple imputation and spline smoothing methods to handle missing data.

RESULTS

When missing data depended on the stochastic progression of disease and medical practice patterns, the spline smoothing method produced results that were close to those obtained when there were no missing data. Compared to multiple imputation, the spline smoothing generally performed similarly or better, with smaller estimation bias and less power loss. The multiple imputation can still reduce study bias and power loss in some restrictive scenarios, eg, when missing data did not depend on the stochastic process of disease progression.

DISCUSSION AND CONCLUSION

Missing data in EHRs could lead to biased estimates of treatment effects and false negative findings in CER even after missing data were imputed. It is important to leverage the temporal information of disease trajectory to impute missing values when using EHRs as a data resource for CER and to consider the missing rate and the effect size when choosing an imputation method.

Minimum Threshold of Bariatric Surgical Weight Loss for Initial Diabetes Remission

OBJECTIVE

There are few studies testing the amount of weight loss necessary to achieve initial remission of type 2 diabetes mellitus (T2DM) following bariatric surgery and no published studies with use of weight loss to predict initial T2DM remission in sleeve gastrectomy (SG) patients.

RESEARCH DESIGN AND METHODS

With Cox proportional hazards models we examined the relationship between initial T2DM remission and percent total weight loss (%TWL) after bariatric surgery. Categories of %TWL were included in the model as time-varying covariates.

RESULTS

Of patients (N = 5,928), 73% were female; mean age was 49.8 ± 10.3 years and BMI 43.8 ± 6.92 kg/m2, and 57% had Roux-en-Y gastric bypass (RYGB). Over an average follow-up of 5.9 years, 71% of patients experienced initial remission of T2DM (mean time to remission 1.0 year). With 0-5% TWL used as the reference group in Cox proportional hazards models, patients were more likely to remit with each 5% increase in TWL until 20% TWL (hazard ratio range 1.97-2.92). When categories >25% TWL were examined, all patients had a likelihood of initial remission similar to that of 20-25% TWL. Patients who achieved >20% TWL were more likely to achieve initial T2DM remission than patients with 0-5% TWL, even if they were using insulin at the time of surgery.

CONCLUSIONS

Weight loss after bariatric surgery is strongly associated with initial T2DM remission; however, above a threshold of 20% TWL, rates of initial T2DM remission did not increase substantially. Achieving this threshold is also associated with initial remission even in patients who traditionally experience lower rates of remission, such as patients taking insulin.

Evaluation of Intensive Telephonic Nutritional and Lifestyle Counseling to Enhance Outcomes of Bariatric Surgery

BACKGROUND

To determine the impact of an intensive perioperative nutritional and lifestyle support protocol on long-term outcomes of bariatric surgery.

METHODS

A retrospective observational study was conducted of 955 patients who underwent gastric bypass surgery between 2005 and 2015. Patients were divided into two cohorts: (1) 2005 through August 2013: these 767 patients were required to participate in the intensive telephone-based nutritional support program from 8 weeks preoperative through 44 weeks postoperative; (2) after August 2013, the program was discontinued and 188 patients did not have intensive telephonic nutritional support. Inverse probability weighting was used to obtain weight loss estimates at 1 and 3 years postoperative. Time-to-event analyses were used to investigate hospitalization rates postoperative. Poisson models were used to investigate healthcare utilization.

RESULTS

Patients who participated in the program exhibited 1.97% (95% CI 0.7, 3.3) greater %TWL at 1 year and 2.2% (95% CI -0.3, 4.1) greater %TWL at 3 years postoperative than patients who did not participate. Secondary analyses indicated participation in the program was associated with 44% shorter time to first hospitalization postoperative (p < 0.001).

CONCLUSIONS

In this health system, intensive nutritional support was associated with greater weight loss at 1 and 3 years postoperative and higher hospitalization rates.

Multiple nutritional deficiencies among adolescents undergoing bariatric surgery: who is at risk?

BACKGROUND

Nutritional deficiencies among adolescents undergoing bariatric surgery (BS) have not been evaluated a in relation to patient's sex.

OBJECTIVES

We compared the preoperative nutritional profile of adolescents characterized by sex and single versus multiple deficiencies.

SETTING

University hospital.

METHODS

Cross-sectional retrospective chart review of 415 eligible adolescents who underwent primary BS between 2011 and 2020. Data included preoperative demographic, anthropometric information as well as three sets of nutritional variables: anemia-related, calcium-related, and other nutritional variables.

RESULTS

The sample comprised 247 males (59.5%) with a mean age of 15.89 ± 1.03 years and a mean body mass index (BMI) of 47.80 ± 6.57 kg/m². Most common deficiencies were vitamin D (92.3%), albumin (51.8%), anemia (15.9%), zinc (11.1%), and vitamin B₁₂ (8%); 21.7% had hyperparathyroidism. Females exhibited a significantly higher prevalence of low hemoglobin, low hematocrit, and iron deficiency. Multiple deficiencies were present among 97.6%, 73.2%, 23.6%, 15%, and 12.6% of adolescents, who had vitamin D, albumin, hemoglobin, zinc, and vitamin B₁₂ deficiencies, respectively. Univariate analysis revealed that adolescents with a BMI of ≥50 kg/m² were 1.24 times more likely to have multiple deficiencies (P = .004). Using multivariate log-binomial regression, BMI of ≥50 kg/m² was a significant predictor of multiple nutritional deficiencies (P = .005, adjusted risk ratio = 1.23, 95% CI 1.06-1.42). Age and sex were not independent predictors of multiple nutritional deficiencies.

CONCLUSION

To our knowledge, this study is the first to appraise single and multiple nutritional deficiencies in adolescents undergoing BS by sex. Multiple deficiencies were common. Females are at higher risk of anemia-related deficiencies. A BMI of ≥50 kg/m² independently and significantly predicted multiple nutritional deficiencies. Correction before and monitoring after surgery are important.