An off-label prescription is one written outside the terms of the marketing authorisation, whether in dose, age, indication or route, while an unlicensed medicine has no authorisation at all in the country where it is given [1]. Neither is unlawful, and in pediatrics neither is unusual. Systematic reviews of hospitalized children have placed off-label prescribing between 12% and 71% of prescriptions, and unlicensed use between 0.2% and 48%, with the youngest patients most heavily exposed [2,3]. The reasons are structural rather than careless: clinical trials in children are difficult to run, so licensing data lag behind practice, and the formulations that reach the market are often designed for adults.

Children are also pharmacologically distinct. Absorption, distribution, metabolism and elimination all shift with age, so a dose extrapolated from adult data may be either ineffective or toxic [4]. Against that background, adverse drug reactions are common. A meta-analysis of prospective studies put the incidence among hospitalized children at close to one in ten [5], a figure consistent with the prospective ward-based work of Martínez-Mir and colleagues [30], and a later systematic review confirmed both the scale of the problem and the inconsistency of the methods used to measure it [6]. Indian data sit within the same range, with reported incidences varying widely by setting and detection method [7,8].

The link between these two observations has been tested directly only a handful of times. Turner and colleagues and Neubert and colleagues both found that unlicensed and off-label medicines were disproportionately implicated in reactions on pediatric wards [9,10]. The most rigorous evidence comes from a nested case-control study by Bellis and colleagues, who analysed 10,699 medicine courses and calculated an odds ratio of 2.25 for an off-label or unlicensed medicine being implicated in a reaction compared with an authorised one [11]. Work on risk factors more broadly has pointed to the number of drugs given, length of stay and particular drug classes as important contributors [12,13].

What is still thin is evidence from Indian district-level teaching hospitals, where prescribing conditions, drug availability and monitoring capacity differ from the tertiary European centres that produced most of this literature. One Indian study has examined off-label use as a risk factor for reactions in children and found an association [14], but replication is scarce. RJDM Medical College & Hospital serves a largely rural population in the Turki area of Bihar. We designed this case-control study to describe our own off-label prescribing, to test whether that exposure carries an independent risk of adverse drug reactions, and to identify which other factors predict them.

REVIEW OF LITERATURE

Early work established that off-label and unlicensed medicines are implicated in reactions more often than their share of prescribing would predict. Turner and colleagues, studying five pediatric wards, found that reactions were significantly more frequent with unlicensed or off-label medicines than with licensed ones [9]. Neubert and colleagues reached a similar conclusion in a German cohort and argued that the association reflected the absence of tested dosing information rather than any property of the drugs themselves [10]. Reviews by Choonara and Conroy, and later by Mason and colleagues, drew these strands together while cautioning that study designs and definitions varied so much that pooled estimates were unreliable [15,16]. Cuzzolin and colleagues made the same point across neonatal and pediatric settings, and Moulis and colleagues have since updated the picture for the pediatric population as a whole [26,29].

The Liverpool programme addressed that criticism with two large linked studies. Thiesen and colleagues followed 6,601 admissions prospectively and characterised the incidence and risk factors of reactions in hospitalized children [17]. Bellis and colleagues then nested a case-control analysis inside that cohort, and it remains the strongest single piece of evidence on this question: the odds of an off-label or unlicensed medicine being implicated in a probable or definite reaction were more than twice those of an authorised medicine, and the hazard rose with each additional drug administered, whether licensed or not [11]. A companion analysis of unplanned admissions found the same pattern among children arriving at hospital because of a reaction [18]. The finding that authorised medicines also raised the hazard is important, because it suggests that prescribing volume and off-label status act partly independently and partly together.

Risk-factor studies outside this programme have converged on similar predictors. Rashed and colleagues, in an international multicentre study, identified the number of medicines and exposure to particular high-risk classes, including antibacterials, anti-epileptics and systemic corticosteroids, as consistent contributors [12]. Indian studies have described the pattern of reactions in hospitalized children, their causality and severity profiles, and the substantial proportion judged preventable [7,19,20]. Saiyed and colleagues examined off-label use specifically as a risk factor in an Indian tertiary hospital and reported a positive association [14]. Cross-sectional surveys from Australia and Iran confirm that off-label and unlicensed prescribing remains widespread wherever it is measured [21,22], while Aronson and Ferner have argued that much of the apparent disagreement between studies is terminological rather than real [1]. Our study contributes a matched case-control analysis from a setting that this literature has largely not reached.

OBJECTIVES

Primary objective:

• To determine whether off-label drug prescribing is associated with an increased risk of adverse drug reactions among hospitalized pediatric patients.

Secondary objectives:

• To identify the pattern and categories of off-label drug use among pediatric inpatients.
• To describe the causality, severity, preventability, organ systems involved and drug classes implicated in the reactions observed.
• To identify independent predictors of adverse drug reactions in hospitalized children.

METHODOLOGY

Study design and setting

This was a hospital-based, matched case-control study conducted by the Department of Pharmacology together with the Department of Pediatrics at RJDM Medical College & Hospital, Turki, Bihar, over twelve months from July 2024 to June 2025. Admissions to the pediatric ward were screened for adverse drug reactions throughout the study period.

Definitions

An adverse drug reaction was defined, following the World Health Organization, as a response to a medicine that is noxious and unintended and occurs at doses normally used in humans for prophylaxis, diagnosis or therapy. A drug was classified as off-label when it was prescribed outside the terms of its marketing authorisation with respect to dose, frequency, age, indication or route, and as unlicensed when no authorisation existed for that product in India [1]. Off-label status was assigned by reference to the approved product literature and the National Formulary of India, independently by two investigators, with disagreements resolved by a third.

Selection of cases and controls

Cases were children aged up to 18 years, admitted to the pediatric ward, who received at least one medicine and developed one or more adverse drug reactions during the admission. Suspected reactions were identified by daily review of case records, drug charts and laboratory results, supplemented by reports from treating clinicians and nursing staff. Only reactions assessed as possible, probable or definite on the Naranjo algorithm were accepted as cases.

Controls were children admitted to the same ward during the same period who received at least one medicine and did not develop any adverse drug reaction. Each case was matched with two controls for age band, sex and primary admitting diagnosis. Matching on these three variables removes them as sources of confounding, but it also means they cannot themselves be assessed as risk factors in this analysis, a point returned to in the limitations.

Sample size

The sample size was calculated for an unmatched case-control comparison with a control-to-case ratio of two. Assuming that 40% of controls would have received at least one off-label drug, and setting the smallest odds ratio worth detecting at 2.0 with 80% power and a two-sided significance level of 5%, the minimum requirement was 99 cases and 198 controls. Allowing for incomplete records, 120 cases and 240 matched controls were enrolled.

Assessment of reactions

Causality was assessed with the Naranjo algorithm, which classifies a reaction as doubtful, possible, probable or definite from a ten-item score [23]. Severity was graded with the modified Hartwig and Siegel scale [24] and preventability with the modified Schumock and Thornton criteria [25]. Reactions were coded by the organ system affected and by the class of the suspected drug.

Statistical analysis

Data were entered in Microsoft Excel and analysed with SPSS version 25. Categorical variables are presented as frequencies and percentages. Because cases and controls were matched, the primary analysis used conditional logistic regression, with crude odds ratios and 95% confidence intervals reported first and then adjusted estimates from a multivariable model. Variables entered into the model were off-label exposure, polypharmacy defined as five or more drugs, receipt of two or more antimicrobials, length of hospital stay beyond seven days, presence of a comorbidity and receipt of three or more parenteral drugs. A dose-response analysis was carried out by grouping patients according to the number of off-label drugs received. A two-sided p-value below 0.05 was treated as significant.

Ethical considerations

The study was approved by the Institutional Ethics Committee of RJDM Medical College & Hospital before enrolment began. Written informed consent was obtained from a parent or legal guardian, with assent from older children where appropriate. All records were anonymised at extraction. Reactions identified during the study were reported to the treating team and, where applicable, to the Pharmacovigilance Programme of India.

Inclusion and Exclusion Criteria

Inclusion criteria:

• Children aged up to 18 years admitted to the pediatric ward during the study period.
• Receipt of at least one medicine during the admission.
• For cases, at least one adverse drug reaction rated possible, probable or definite on the Naranjo algorithm.
• Consent from a parent or legal guardian.

Exclusion criteria:

• Reactions classified as doubtful on causality assessment, and events attributable to the underlying disease, drug overdose, poisoning or deliberate self-harm.
• Neonates admitted to the neonatal intensive care unit, whose prescribing and monitoring follow separate protocols.
• Admissions of less than 24 hours, and children who left against medical advice before assessment was complete.
• Incomplete case records, and patients for whom no suitable matched control could be identified.

RESULTS AND ANALYSIS

A total of 120 cases and 240 matched controls were analysed. Because the groups were matched for age, sex and primary diagnosis, these characteristics were closely comparable by design, as shown in Table 1. The two groups differed in the number of medicines received and in length of stay.

Table 1. Baseline characteristics of cases and controls. Age, sex and primary diagnosis were matching variables and are therefore comparable by design.

Across the 360 children, 1,584 drugs were prescribed, of which 612 (38.6%) were off-label or unlicensed. Dose or frequency outside the licensed range was the commonest reason, followed by use in a child below the licensed age. The distribution of categories is given in Table 2.

Table 2. Pattern of off-label and unlicensed prescribing (n = 1,584 drugs prescribed; 612 off-label, 38.6%).

The 120 cases experienced 138 adverse drug reactions. Most were rated probable on causality assessment, most were mild or moderate in severity, and half were judged preventable. These characteristics appear in Table 3.

Table 3. Causality, severity and preventability of the 138 adverse drug reactions recorded in 120 cases.

Gastrointestinal and cutaneous reactions predominated, and antimicrobials were the drug class implicated most often. Table 4 sets out the organ systems affected and the classes responsible.

Table 4. Organ systems affected and drug classes implicated in the 138 reactions.

On univariate analysis, every exposure examined was associated with adverse drug reactions. At least one off-label drug had been received by 76.7% of cases against 55.0% of controls, giving a crude odds ratio of 2.69. Polypharmacy carried the largest crude effect. The unadjusted results are given in Table 5.

Table 5. Univariate analysis of exposures associated with adverse drug reactions.

After adjustment in the multivariable conditional logistic regression model, four exposures retained independent significance: off-label drug exposure, polypharmacy, prolonged hospital stay and receipt of two or more antimicrobials. Comorbidity and parenteral drug burden lost significance once the other variables were accounted for. The adjusted estimates are presented in Table 6 and displayed in Figure 1.

Table 6. Independent predictors of adverse drug reactions on multivariable conditional logistic regression. Age, sex and primary diagnosis were matching variables and were not entered as candidate predictors.

Figure 1. Forest plot of adjusted odds ratios for adverse drug reactions. Estimates to the right of the dashed line indicate increased risk; red markers denote statistically significant predictors.

The relationship with off-label exposure was graded rather than all-or-none. Among cases, 28 children (23.3%) had received no off-label drug, 48 (40.0%) had received one or two and 44 (36.7%) had received three or more; the corresponding figures among controls were 108 (45.0%), 96 (40.0%) and 36 (15.0%). Taking children with no off-label exposure as the reference, the adjusted odds ratio was 1.68 (95% CI 0.96 to 2.94) for one or two off-label drugs and 3.12 (95% CI 1.72 to 5.66) for three or more, a trend that was significant (p for trend < 0.001). This gradient is shown in Figure 2.

Figure 2. Distribution of cases and controls by number of off-label drugs received, showing a higher off-label burden among children who developed a reaction.

Of the 13 severe reactions, nine occurred in children who had received at least one off-label drug. No reaction resulted in death, and all severe reactions resolved after withdrawal of the suspected drug and supportive treatment.

DISCUSSION AND INTERPRETATION

Three findings stand out. Off-label prescribing was common, affecting nearly two-fifths of all drugs and the majority of children. It carried an independent association with adverse drug reactions that survived adjustment for prescribing volume, antimicrobial burden and length of stay. And the association strengthened as the number of off-label drugs increased, which is the pattern one expects when an exposure is contributing to an outcome rather than merely accompanying it.

Our adjusted odds ratio of 2.31 sits close to the 2.25 reported by Bellis and colleagues in their nested case-control study, which remains the reference point for this question [11]. That two very different settings, a specialist children's hospital in England and a district teaching hospital in Bihar, should produce almost the same estimate is reassuring for the finding and suggests the mechanism is not local. The most plausible explanation is the one Neubert and colleagues advanced: when a medicine is given outside its licence, the dosing and safety information that would normally guide its use has not been generated for that age group or indication, so the margin for error is wider [10]. Turner and colleagues reached the same conclusion from ward-based data two decades ago [9]. The categories we found support this reading, since dose or frequency deviations and use below the licensed age together accounted for almost two-thirds of off-label prescribing, and both are situations where the prescriber is extrapolating rather than following tested guidance [4].

Polypharmacy carried an effect at least as large as off-label status, with an adjusted odds ratio of 2.48 for five or more drugs. This matches the international multicentre findings of Rashed and colleagues, who identified the number of medicines as a consistent predictor across countries [12], and the Liverpool cohort, where the hazard rose with each additional drug whether or not it was licensed [11,17]. The practical implication is that off-label prescribing and polypharmacy are partly separate problems: reducing one will not automatically fix the other, and a child on five authorised drugs is not obviously safer than a child on three off-label ones. The antimicrobial finding fits the same literature, with antibacterials repeatedly identified as a high-risk class [12], and in our series antimicrobials accounted for 44% of implicated drugs, in a ward where infections drove most admissions. Prolonged stay is best read as a marker of severity and of cumulative exposure rather than as a cause in its own right.

The profile of the reactions themselves was unremarkable and broadly matches Indian and international series. Gastrointestinal and cutaneous reactions predominated, most reactions were mild or moderate, and probable causality was the commonest rating, as reported by Kurian and colleagues, Gupta and colleagues, and Digra and colleagues in Indian pediatric wards [7,19,20]. The figure that deserves attention is preventability: half the reactions met the Schumock and Thornton criteria for being definitely or probably preventable. Preventable harm is, by definition, harm that a system change could avoid, and taken together with the dose-related nature of much off-label prescribing this points to specific interventions. A weight-based dosing chart validated against the National Formulary of India, a pharmacist check on prescriptions for children receiving five or more drugs, and active surveillance rather than spontaneous reporting for children on multiple off-label medicines would each address a mechanism our data implicate. Strengthening reporting to the Pharmacovigilance Programme of India would also improve the evidence base, since Indian pediatric reporting remains sparse relative to the burden [8].

Two cautions temper the interpretation. First, an association of this kind cannot establish causation: children who receive off-label drugs may be sicker in ways that matching on diagnosis does not fully capture, and residual confounding by severity is plausible even after adjustment. Second, off-label status is a heterogeneous label covering everything from a modest dose deviation to use of an adult formulation in an infant, and Aronson and Ferner have argued convincingly that treating it as a single exposure obscures more than it reveals [1]. Our category analysis is a partial answer, but a larger study would be needed to estimate risk separately for each category. It also bears repeating that off-label prescribing is frequently the right clinical decision rather than an avoidable error, a position professional bodies have set out explicitly [27]; the aim is safer off-label prescribing, not its elimination. Surveys suggest that prescriber awareness of off-label status is itself uneven [28], so simply making that status visible on the drug chart may be a useful first step. What our findings do establish is that off-label exposure marks out a group of hospitalized children at measurably higher risk, and that this group can be identified at the bedside without any additional test.

CONCLUSION

Off-label prescribing was widespread among children admitted to this hospital, involving 38.6% of all drugs and more than three-quarters of those who developed an adverse drug reaction. After adjustment for prescribing volume, antimicrobial exposure and length of stay, off-label exposure remained independently associated with a roughly two-fold increase in the odds of a reaction, and the risk climbed further in children receiving three or more off-label drugs. Polypharmacy, prolonged hospital stay and receipt of two or more antimicrobials were also independent predictors. Because half the reactions were judged preventable, these findings translate directly into practice: children receiving multiple or off-label medicines should be flagged for active monitoring, weight-based dosing should be checked against a standard reference before administration, and antimicrobial prescribing should be reviewed regularly. For a teaching hospital serving rural Bihar, these are achievable steps that would reduce avoidable harm without requiring new resources.

Limitations of the Study

• A case-control design shows association rather than causation, and residual confounding by underlying illness severity cannot be excluded despite matching and adjustment.
• Age, sex and primary diagnosis were used as matching variables, so their contribution as risk factors could not be evaluated in this analysis.
• Detection of reactions relied on record review and clinician reporting, so mild, transient or delayed reactions were probably missed, and misclassification of some controls is possible.
• Off-label status was treated as a single exposure although the categories differ substantially in the degree of extrapolation involved, and the study was not powered to estimate risk within each category.
• This was a single-centre study, which limits generalisability, and the sample was too small to examine rare or severe reactions in any detail. Multi-centre prospective studies with active surveillance would give more precise and more transferable estimates.

REFERENCES

1. Aronson JK, Ferner RE. Unlicensed and off-label uses of medicines: definitions and clarification of terminology. Br J Clin Pharmacol. 2017;83(12):2615-2625.
2. Magalhães J, Rodrigues AT, Roque F, Figueiras A, Falcão A, Herdeiro MT. Use of off-label and unlicenced drugs in hospitalised paediatric patients: a systematic review. Eur J Clin Pharmacol. 2015;71(1):1-13.
3. Gore R, Chugh PK, Tripathi CD, Lhamo Y, Gautam S. Pediatric off-label and unlicensed drug use and its implications. Curr Clin Pharmacol. 2017;12(1):18-25.
4. Kearns GL, Abdel-Rahman SM, Alander SW, Blowey DL, Leeder JS, Kauffman RE. Developmental pharmacology: drug disposition, action, and therapy in infants and children. N Engl J Med. 2003;349(12):1157-1167.
5. Impicciatore P, Choonara I, Clarkson A, Provasi D, Pandolfini C, Bonati M. Incidence of adverse drug reactions in paediatric in/out-patients: a systematic review and meta-analysis of prospective studies. Br J Clin Pharmacol. 2001;52(1):77-83.
6. Smyth RMD, Gargon E, Kirkham J, Cresswell L, Golder S, Smyth R, et al. Adverse drug reactions in children: a systematic review. PLoS One. 2012;7(3):e24061.
7. Kurian J, Mathew J, Sowjanya K, Chaitanya KRK, Ramesh M, Sebastian J, et al. Adverse drug reactions in hospitalized pediatric patients: a prospective observational study. Indian J Pediatr. 2016;83(5):414-419.
8. Patel TK, Patel PB. Incidence of adverse drug reactions in Indian hospitals: a systematic review of prospective studies. Curr Drug Saf. 2016;11(2):128-136.
9. Turner S, Nunn AJ, Fielding K, Choonara I. Adverse drug reactions to unlicensed and off-label drugs on paediatric wards: a prospective study. Acta Paediatr. 1999;88(9):965-968.
10. Neubert A, Dormann H, Weiss J, Egger T, Criegee-Rieck M, Rascher W, et al. The impact of unlicensed and off-label drug use on adverse drug reactions in paediatric patients. Drug Saf. 2004;27(13):1059-1067.
11. Bellis JR, Kirkham JJ, Thiesen S, Conroy EJ, Bracken LE, Mannix HL, et al. Adverse drug reactions and off-label and unlicensed medicines in children: a nested case-control study of inpatients in a pediatric hospital. BMC Med. 2013;11:238.
12. Rashed AN, Wong ICK, Cranswick N, Tomlin S, Rascher W, Neubert A. Risk factors associated with adverse drug reactions in hospitalised children: international multicentre study. Eur J Clin Pharmacol. 2012;68(5):801-810.
13. Le J, Nguyen T, Law AV, Hodding J. Adverse drug reactions among children over a 10-year period. Pediatrics. 2006;118(2):555-562.
14. Saiyed MM, Lalwani T, Rana D. Is off-label use a risk factor for adverse drug reactions in pediatric patients? A prospective study in an Indian tertiary care hospital. Int J Risk Saf Med. 2015;27(1):45-53.
15. Choonara I, Conroy S. Unlicensed and off-label drug use in children: implications for safety. Drug Saf. 2002;25(1):1-5.
16. Mason J, Pirmohamed M, Nunn T. Off-label and unlicensed medicine use and adverse drug reactions in children: a narrative review of the literature. Eur J Clin Pharmacol. 2012;68(1):21-28.
17. Thiesen S, Conroy EJ, Bellis JR, Bracken LE, Mannix HL, Bird KA, et al. Incidence, characteristics and risk factors of adverse drug reactions in hospitalised children: a prospective observational cohort study of 6,601 admissions. BMC Med. 2013;11:237.
18. Bellis JR, Kirkham JJ, Nunn AJ, Pirmohamed M. Adverse drug reactions and off-label and unlicensed medicines in children: a prospective cohort study of unplanned admissions to a paediatric hospital. Br J Clin Pharmacol. 2014;77(3):545-553.
19. Gupta S, Zaki SA, Masavkar S, Shanbag P. Causality, severity, and avoidability of adverse drug reactions in hospitalized children: a prospective cohort study. Cureus. 2023;15(1):e33369.
20. Digra KK, Pandita A, Saini GS, Bharti R. Pattern of adverse drug reactions in children attending the department of pediatrics in a tertiary care center: a prospective observational study. Clin Med Insights Pediatr. 2015;9:73-78.
21. Landwehr C, Richardson J, Bint L, Parsons R, Sunderland B, Czarniak P. Cross-sectional survey of off-label and unlicensed prescribing for inpatients at a paediatric teaching hospital in Western Australia. PLoS One. 2019;14(1):e0210237.
22. Kouti L, Aletayeb M, Aletayeb SMH, Hardani AK, Eslami K. Pattern and extent of off-label and unlicensed drug use in neonatal intensive care units in Iran. BMC Pediatr. 2019;19(1):3.
23. Naranjo CA, Busto U, Sellers EM, Sandor P, Ruiz I, Roberts EA, et al. A method for estimating the probability of adverse drug reactions. Clin Pharmacol Ther. 1981;30(2):239-245.
24. Hartwig SC, Siegel J, Schneider PJ. Preventability and severity assessment in reporting adverse drug reactions. Am J Hosp Pharm. 1992;49(9):2229-2232.
25. Schumock GT, Thornton JP. Focusing on the preventability of adverse drug reactions. Hosp Pharm. 1992;27(6):538.
26. Cuzzolin L, Atzei A, Fanos V. Off-label and unlicensed prescribing for newborns and children in different settings: a review of the literature and a consideration about drug safety. Expert Opin Drug Saf. 2006;5(5):703-718.
27. Frattarelli DA, Galinkin JL, Green TP, Johnson TD, Neville KA, Paul IM, et al. Off-label use of drugs in children. Pediatrics. 2014;133(3):563-567.
28. Balan S, Hassali MA, Mak VSL. Awareness, knowledge and views of off-label prescribing in children: a systematic review. Br J Clin Pharmacol. 2015;80(6):1269-1280.
29. Moulis F, Durrieu G, Lapeyre-Mestre M. Off-label and unlicensed drug use in children population. Therapie. 2018;73(2):135-149.
30. Martínez-Mir I, García-López M, Palop V, Ferrer JM, Rubio E, Morales-Olivas FJ. A prospective study of adverse drug reactions in hospitalized children. Br J Clin Pharmacol. 1999;47(6):681-688.