Button Battery Ingestion in Infants: A Systematic Review

Button battery exposures among children pose a serious public health problem. During the last 40 years, button battery ingestion (BBI) frequency and outcome severity have increased, with moderate and fatal cases rising over 18-fold from 0.4% in 1985 to 7.5% in 2019 [1]. Toddlers account for most BBI [2]. Risks are exacerbated by the rapid onset of tissue damage within hours, diagnostic delays from unwitnessed ingestions and variable symptoms at presentation in non or pre-verbal children [2]. Recently, Smetak and Wilcox reported BBI in an 11-month-old who presented to the emergency department with a two-week history of dysphagia and cough [3]. A chest X-ray revealed a corroded button battery (BB) in the proximal oesophagus. Ten hours post-removal, the infant developed haematemesis and shock and intraoperative angiography showed an oesophageal fistula and perforation of the left common carotid artery. The child survived and was discharged after 32 days [3]. Our recent systematic literature review and pooled analysis of individual patient-level data identified and described complications in 399 cases of BBI [2]. Publication of the case report [3] prompted additional analyses in the subgroup of children aged < 1 year to summarise the impact of BBI in infants, present key clinical characteristics (symptom progression and complications), and compare outcomes to those in older children (≥ 1 year and < 18 years of age). We followed PRISMA 2020 guidelines and systematically searched MEDLINE, Embase and Scopus databases up to May 19, 2023. Search terms included “child”/“infant”/“adolescent” and “button battery”/“disc battery” and “ingestion”/“insertion”/“inhalation” and “complications”/“injuries” [2]. We included studies of any design that described BBI complications or injuries in children and provided individual patient-level data. References of relevant articles were searched for eligible studies, but unpublished literature was not. Data from relevant studies were collected and managed using REDCap and included information on author, publication year, study type, participants' demographics, exposure characteristics, symptoms at presentation and details of injury, treatment, and complications. Symptoms were categorised as pain-related, feeding/swallowing difficulties, upper gastrointestinal (GI) upset (including nausea, vomiting, or retching), lower GI upset (diarrhoea, abdominal distension, constipation, or intestinal obstruction), upper airway symptoms, respiratory difficulties, bleeding, systemic and other [2]. Adjusted odds ratios (OR) with 95% confidence intervals (CIs) were calculated using binary logistic regression to compare BBI in infants (aged < 1 year) and older children (≥ 1 year and < 18 years), using IBM SPSS, version 25.0. An alpha below 0.05 was considered significant. Of 2031 articles identified, 204 studies with 399 children met our inclusion criteria: 146 (37%) infants and 253 (63%) children ≥ 1 year age [2]. New analyses of 146 infants showed a mean age at the time of BBI of 0.75 years SD 0.38. Among 127 with sex reported, males accounted for 53%. Of 130 cases, 97% ingested only one BB. Battery size was reported in 82 cases, being ≥ 20 mm diameter in 61 (74%). BBs were removed in ≥ 6 h in 60%. One infant also ingested parts of a fidget spinner. BBI was witnessed in only a fifth of exposures and the most common sites of impaction were the oesophagus (77%; proximal oesophagus 65%, mid 17%, distal 4% and unspecified 14%) and the stomach (8%). Sex-adjusted analyses revealed infants had higher rates of feeding difficulties (aOR 1.85, 95% CI: 1.19–2.89), p = 0.006, upper airway difficulties (aOR: 4.10, 95% CI: 2.21–7.61), p < 0.001 and respiratory difficulties (aOR: 2.17, 95% CI: 1.38-3.40), p = 0.001 at presentation compared to older children (Figure 1). At 6 h post-ingestion, infants mainly presented with feeding/swallowing difficulties (58%), and respiratory difficulties (38%). Beyond 6 h, main presenting symptoms included feeding difficulties (60%), respiratory difficulties (46%) and systemic symptoms (44%). Infants were more likely than older children to develop vocal cord paralysis (aOR: 2.77, 95% CI: 1.28–6.02), p = 0.01, tracheooesophageal fistula (aOR: 2.97, 95% CI: 1.79–4.95), p < 0.001, and to die (aOR: 2.23, 95% CI: 1.08–4.61), p = 0.03 (Figure 1). This study highlights challenges distinguishing BBI in infants on initial presentation due to non-specific symptoms and unwitnessed ingestions, necessitating high clinical suspicion. When suspected, abdominal and chest X-rays are required including anteroposterior and lateral projections [4] since BB injuries are time-critical and may cause severe and life-threatening complications [2, 4]. Our data show infants with BBI at greater risk of vocal cord paralysis, tracheooesophageal fistula and death than older children. A sustained, coordinated, multifaceted preventive effort involving health professionals, manufacturers, policymakers, and the public is required to protect children from BBI harms. This includes mandatory legislation for tool-assisted child-resistant closures on all BB-containing products and compliance testing and penalties for non-compliance. Industry involvement should enhance product durability with child-resistant packaging and improve battery compartment design [5]. Tailored educational programs for health professionals, parents and caregivers of young children would be impactful given the disproportionate risks [2]. Strategies to reduce inadvertent BBI, including alternative power sources, merit consideration. BB redesigns with pressure-sensitive coatings to prevent conduction, or built-in fuses to stop current emission post-ingestion, should be prioritised to prevent tissue damage [4]. Carlos Nunez: conceptualization, formal analysis, writing – original draft, methodology, data curation. Elizabeth J. Elliott: supervision, validation, writing – review and editing, formal analysis, data curation. The APSU is funded by the Queensland Department of Health (Ref ID: PSQIS 8309) to conduct surveillance for battery exposures in children. The Australian Government Department of Health and Aged Care also funds the APSU (Health/21-22/D21-5425703). E.J.E. is supported by a NHMRC Investigator Grant (#2026176). The authors declare no conflicts of interest. The data that support the findings of this study are available from the corresponding author upon reasonable request.