Inverse-probability weighting and multiple imputation for evaluating selection bias in the estimation of childhood obesity prevalence using data from electronic health records

Abstract

Background and objectives: Height and weight data from electronic health records are increasingly being used to estimate the prevalence of childhood obesity. Here, we aim to assess the selection bias due to missing weight and height data from electronic health records in children older than five. Methods: Cohort study of 10,811 children born in Navarra (Spain) between 2002 and 2003, who were still living in this region by December 2016. We examined the differences between measured and non-measured children older than 5 years considering weight-associated variables (sex, rural or urban residence, family income and weight status at 2-5 yrs). These variables were used to calculate stabilized weights for inverse-probability weighting and to conduct multiple imputation for the missing data. We calculated complete data prevalence and adjusted prevalence considering the missing data using inverse-probability weighting and multiple imputation for ages 6 to 14 and group ages 6 to 9 and 10 to 14. Results: For 6-9 years, complete data, inverse-probability weighting and multiple imputation obesity age-adjusted prevalence were 13.

Completeness and representativeness of body mass index in children's electronic general practice records: Linked cross‐sectional study in an ethnically‐diverse urban population in the United Kingdom

“…We found children living in more deprived areas were more likely to ever have a GP‐BMI record, in contrast to findings from a study in Spain where the prevalence of child height and weight records in GP‐EHRs increased with average annual household income 27 …”

Risk for cardiovascular disease associated with metabolic syndrome and its components: a 13-year prospective study in the RIVANA cohort

“…Despite the aforementioned limitations, the strengths of this study rely on the population-based nature of the cohort that includes a representative population between 35 and 84 years, the long follow-up (median 12.8 years, interquartile range 12.5 to 13.1), very high retention rate (> 99%), the rigorous method of collection of cardiovascular events supported by public resources, the use of an electronic-health records database that has been previously validated with other health conditions [ 31 , 32 ], and the inclusion of sensitivity analyses with specific waist circumference cutoff levels for Spanish population [ 44 ].…”

“…This database was designed to account for diverse episodes and health conditions, and was coded according to the International Classification of Primary Care, Second Edition (ICPC-2) [ 30 ]. The database has been previously validated with other disease diagnoses such as diabetes type 2 and childhood obesity, resulting in a valid source for epidemiological surveillance [ 31 , 32 ]. Moreover, in Spain, with a universal healthcare system, primary health care is frequently used by the Spanish population (in 2016, the frequency of medical consultations was 5.2 visits per person/year [ 33 ]).…”

Small single perivascular hepatocellular carcinoma: comparisons of radiofrequency ablation and microwave ablation by using propensity score analysis

“…In this study, we compared the effectiveness and survival outcomes of these two ablation therapies. Given the potential selection bias in treatment assignment, we adopted the inverse probability of treatment weighting (IPTW) based on the propensity to receive treatment to reduce these bias [15].…”