Gastroesophageal reflux (GER) occurs in up to 2/3 of healthy infants1 and usually resolves by 1 year of life. In contrast to GER, gastroesophageal reflux disease (GERD) is reflux that can involve (a) poor weight gain, (b) pain or (c) mucosal injury on endoscopy.2 Acid suppressants such as histamine-2 receptor antagonists (H2RA) and proton-pump inhibitors (PPI) reduce gastric acidity but are not effective against GER.2–4 Acid suppressants have been linked to a higher risk of serious infections5,6 and fractures.7 Weighing the risks of these medications against their unclear benefit, Choosing Wisely and the American Academy of Pediatrics have recommended against their use in infants with GER. Our objective was to determine the individual and health care system characteristics associated with acid suppressant overuse using a statewide all-payers claims database from the most recent years available with newer International Statistical Classification of Diseases, 10th Revision (ICD-10) codes that can distinguish GER from GERD.
Methods
The data were obtained from the Virginia All-Payers Claim Database. The patient’s ZIP code was classified on the basis of the Education Demographic and Geographic Estimates Program.8 Infants 0 to 11 months of age with at least 30 days of continuous enrollment between 2016 to 2019 were included. We excluded infants who had been diagnosed with a Pediatric Complex Chronic Condition at any point in their patient record.9 We converted other exclusion criteria used in earlier claims-based studies,10 such as esophagitis, ulcers, and weight loss (full list available in Supplemental Table 2) from International Statistical Classification of Diseases, 9th Revision (ICD-9) to ICD-10 codes, and these were applied if they occurred 0 to 1 days before the prescription of a H2RA or PPI. We conducted 2 separate analyses: 1 in which GERD was an exclusion criterium and 1 in which it was not. We modeled the rate of H2RA or PPI use using a multivariable binomial regression model including sex, insurance type, rurality, and indicators for low birth weight and prematurity. All statistical analyses were performed at a 5% statistical significance level by using R (version 4.1.0). Ethical approval was given by the Virginia Commonwealth University institutional review board with an exempt determination.
Results
We identified 270 437 total infants (68% of live births in Virginia from 2016–2019). There were 16 910 (7%) nonmedically complex children who were prescribed a H2RA or PPI (Fig 1). There were 5196 (2%) infants excluded with a diagnosis of GERD. There were 5433 (2%) unique children without one of the exclusion criteria who were prescribed a H2RA or a PPI. The odds of being prescribed H2RA or PPI were higher (adjusted odds ratio [aOR] 1.9; 95% confidence interval [CI] 1.8–2.0) for infants with commercial insurance compared to those with public insurance (Table 1). Children from rural settings had higher odds of being prescribed H2RA or PPI compared to children from both urban (aOR 1.2; 95% CI 1.1–1.3) and suburban (aOR 1.6; 95% CI 1.5–1.7) settings. Removing GERD as an exclusion factor did not meaningfully or substantially alter the estimates (Supplemental Table 3).
Flow diagram of included and excluded infants prescribed an antacid. Numbers reflect GERD as an exclusion criterium.
Flow diagram of included and excluded infants prescribed an antacid. Numbers reflect GERD as an exclusion criterium.
Adjusted Odds Ratios and 95% Confidence Intervals for Infants Prescribed Acid Suppressant Medication Without One of the Exclusion Criteria
| Categorya | aOR (95% CI) | P |
|---|---|---|
| Male | 1.11 (1.05–1.18) | .001b |
| Commercial insurance | 1.86 (1.76–1.97) | <.001b |
| Domicile | ||
| Rural | 1.60 (1.48–1.73) | <.001b |
| City | 1.30 (1.18–1.43) | <.001b |
| Prematurity | 1.16 (1.00–1.35) | .05 |
| Low birth weight | 1.15 (0.95–1.41) | .15 |
| Categorya | aOR (95% CI) | P |
|---|---|---|
| Male | 1.11 (1.05–1.18) | .001b |
| Commercial insurance | 1.86 (1.76–1.97) | <.001b |
| Domicile | ||
| Rural | 1.60 (1.48–1.73) | <.001b |
| City | 1.30 (1.18–1.43) | <.001b |
| Prematurity | 1.16 (1.00–1.35) | .05 |
| Low birth weight | 1.15 (0.95–1.41) | .15 |
Reference categories are female, public insurance, suburban domicile, full term, and non-low-birth weight infants
Indicates result is significant at P < .05.
Discussion
In our statewide analysis, nearly 7% of all infants and 2% of infants without one of the exclusion criteria were prescribed an acid suppressant. Earlier studies found that 2% to 6% of infants were prescribed an acid suppressant,10 although the change from ICD-9 to ICD-10 codes and lack of 1:1 mapping makes comparisons with older data challenging. Our estimate being at the lower end of this range may be a result of the 2014 Choosing Wisely recommendation, the growing recognition of the harms of acid suppressants, or the updated exclusion criteria that we employed. Providers often have difficulty distinguishing GER from GERD. Observations of poor weight gain and mucosal changes on endoscopy are relatively straightforward and free from provider interpretation. However, the assessment of pain is more subjective since an infant may cry for many different reasons and the peak age for reflux overlaps with that of normal developmental crying patterns. Furthermore, some providers may feel compelled to “upcode” from a diagnosis of GER to a diagnosis of GERD to justify their medical management. In conclusion, acid suppressant overuse remains a persistent problem particularly among commercially insured infants and those residing in rural areas. Reduction efforts should include providers and patients outside of urban academic children’s hospitals to achieve maximal benefit.
Dr Wolf conceptualized the study, interpreted the results, drafted, and revised the manuscript; Mr French and Dr Lavallee queried the databases, conducted analyses, and revised the manuscript; Dr Sabo helped design the study, supervised the queries and analyses, helped interpret the results, and revised the manuscript; Drs Schroeder, Huffstetler, Schefft, and Krist helped design the study, interpret the results, and revise the manuscript; and all authors approved the final manuscript as submitted and agree to be accountable for all aspects of the work.
CONFLICT OF INTEREST DISCLOSURES: The authors have no conflicts of interest relevant to this article to disclose.
FUNDING: Dr Wolf was funded through the National Center for Advancing Translational Sciences [KL2TR002648]. This project was supported in part by the Biostatistics, Epidemiology and Research Design (BERD) core. Drs Sabo and Krist and Mr French received salary support through C. Kenneth and Dianne Wright Center for Clinical and Translational Research (CTSA award UL1TR002649) from the National Center for Advancing Translational Sciences. This project was also funded by a Virginia Commonwealth University Center for Clinical and Translational Research Endowment Award. The funders played no role in the work.
Comments