Study Design and Patient Population
This study was based on the data from the JADE study for pediatric inpatients, which was a historical cohort study performed in two tertiary care teaching hospitals in Japan. The details of the study have been described elsewhere [11]. Briefly, we included all patients aged ≤15 years admitted to any ward, including the neonatal intensive care unit (NICU) and pediatric intensive care unit (ICU), and patients aged >15 years admitted to any pediatric ward over a 3-month period in 2009. Because some adult patients with congenital or metabolic diseases were cared for by pediatricians from a young age, such patients were included in this cohort study based on the protocol. We excluded neonates in well-baby nurseries from this study because they were healthy and not cared for by pediatricians. If neonates had a problem such as temporary dyspnea or mild cyanosis of the limbs at birth, they were admitted to the NICU and cared for by neonatologists. We included these neonates in this study. We categorized the age groups as follows: neonates (aged <1 month), infants (1 month to <1 year), preschoolers (1 year to <7 years), school-aged children (7 to <13 years), teenagers (13 to <19 years), and adults (≥19 years).
The institutional review boards of the two participating hospitals approved the study. Because all data were obtained as part of routine daily practice, the institutional review boards waived the need for informed consent.
Definitions
The primary outcome of the study was the occurrence of ADEs, which we compared between pediatric patients with and without cancer. Cancer patients were defined as those who were diagnosed with any malignant tumor or those who had a tumor and were receiving antitumor agents. Non-cancer patients included those with benign or other tumors. We used validated methodology for the classification of ADEs [12]. An ADE was defined as a health injury occurring because of medication use. For example, nausea or vomiting in a patient receiving an antitumor agent was considered an ADE. We categorized the severity of ADEs as follows: fatal (resulting in death), life-threatening (requiring transfer to the ICU or causing anaphylactic shock), serious (neutropenia requiring a special protective environment, cutaneous lesions requiring therapy, gastrointestinal bleeding, altered mental status, excessive sedation, increased creatinine level, or decreased blood pressure), or significant (rash, diarrhea, or nausea). Categories of ADE symptoms included bleeding; central nervous system; allergic or skin reaction; liver or metabolic dysfunction; cardiovascular; gastrointestinal; renal; respiratory; bone marrow suppression or cytopenia; and other.
We categorized medications as follows: antihistamines, antibiotics, antitumor agents, adrenaline/anticholinergics, blood products, hematopoietic drugs, anticoagulants, diuretics/cardiovascular agents, antipyretic analgesics/nonsteroidal anti-inflammatory drugs (NSAIDs), anticonvulsants, sedatives, antipsychotics, diagnostic drugs/electrolytes and fluids/others, antitussives, ophthalmic/otolaryngologic/dermatologic drugs, laxatives, local anesthetics, corticosteroids, hormones/insulin, aminophylline, and peptic ulcer drugs. Antitussives did not include codeine but did include expectorants, and sedatives did not include narcotics or opiates. Because doses for pediatric patients were generally determined by body weight, and the standard doses varied between drugs, we did not account for dose in the analyses.
Data Collection and Review Process
Trained reviewers based at each participating hospital reviewed all medical charts, laboratory results, incident reports, and prescription queries from pharmacists. The trained reviewers included a board-certified pediatrician, pediatric nurses, and a dietitian; the pediatrician trained all reviewers in a standard manner, as reported elsewhere [12]. Reviewers collected the characteristics and administrative data for all patients enrolled in the cohort and identified potential ADEs and associated details, such as detailed symptoms and drug name, dose, route, and class.
After data collection, two independent physician reviewers assessed, in a standard manner, whether any potential ADEs should be classified as ADEs [12]. Briefly, the reviewers summarized and discussed many aspects, including preceding drugs, other causative conditions occurring during hospitalization, previous literature reports, alleviation after discontinuation of drug, repeated symptoms when the same drug was re-introduced, and so on. They classified the severity, symptoms, and class of medication involved in ADEs. When disagreement arose over classification of an event, the reviewers reached consensus through discussion. Uncertain symptoms or those for which consensus was not reached were excluded from the ADEs.
Statistical Analyses
Categorical variables regarding patient characteristics are reported as numbers and percentages. A Chi squared test was used to compare patients with and without cancer. We also constructed a logistic regression model for cancer patients who developed ADEs, adjusting for the age group and admission to an ICU. The likelihood of ADEs was expressed as an odds ratio (OR) and its 95% confidence interval (CI). The ADE rate per 100 patients, ADE severity, and ratio of ADE severity for each drug were compared between cancer and non-cancer patients; the Chi squared test was used for categorical variables.
We carried out all analyses using JMP 12.0 software (SAS Institute Inc., Cary, NC, USA). Two-tailed p values <0.05 were considered statistically significant.