Paediatric off-label use of drugs in Gansu, China: a multicentre cross-sectional study

Abstract

Objective

To examine the current prevalence and cost of paediatric off-label drug prescriptions in Gansu, China, and the potential influencing factors.

Design

The prevalence of off-label prescriptions in paediatrics was evaluated according to the National Medical Products Administration drug instructions in the China Pharmaceutical Reference (China Pharmaceutical Reference, MCDEX) database. The evidence of the prescription was determined by existing clinical practice guidelines and the Thomson Grade in the Micromedex 2021 compendium. We used logistic regression to investigate the characteristics that influence paediatric off-label drug use after single-factor regression analysis.

Setting

A multicentre cross-sectional study of outpatient paediatric prescriptions in 196 secondary and tertiary hospitals in Gansu Province, China, in March and September 2020.

Results

We retrieved 104 029 paediatric prescriptions, of which 39 480 (38.0%) contained off-label use. The most common diseases treated by off-label drugs were respiratory system diseases (n=15 831, 40.1%). A quarter of off-label prescriptions had adequate evidence basis (n=10 130, 25.6%). Unapproved indications were the most common type of off-label drug use (n=25 891, 65.6%). A total of 1177 different drugs were prescribed off-label, with multienzyme tablets being the most common drug (n=1790, 3.5%). The total cost of the prescribed off-label drugs was ¥106 116/day. Off-label prescriptions were less frequent in tertiary than in secondary hospitals. Topical preparations were more commonly prescribed off-label than other types of drugs. Senior-level clinicians prescribed drugs off-label more often than intermediate and junior clinicians.

Conclusion

Off-label drug use is widespread in paediatric practice in China. Three-quarters of the prescriptions may potentially include inappropriate medication use, resulting in a daily economic burden of about ¥81 000 in 2020 in Gansu Province with 25 million inhabitants. The management of off-label drug use in paediatrics in China needs improvement.

STRENGTHS AND LIMITATIONS OF THIS STUDY.

• This study is a multicentre cross-sectional study of outpatient paediatric prescriptions in secondary and tertiary hospitals in Gansu Province, China. To our knowledge, this is so far the largest paediatric outpatient prescription analysis in China.

• Six pharmacists manually evaluated 104 029 prescriptions, spending a great effort to determine the reasonableness of off-label drug use.

• This study estimated the economic burden caused by unnecessary off-label drug use, to investigate the importance of promoting the rational drug use in children.

• We determined the off-label drug based on the instructions of the National Medical Products Administration in the China Pharmaceutical Reference (China Pharmaceutical Reference, MCDEX) database, which may not be completely up to date.

• This study used clinical practice guidelines and the Thomson Grade in the Micromedex 2021 compendium to determine the distribution of evidence-based medicine evidence, which may have some gaps.

Introduction

Off-label drug use is a great challenge for public health in children and adolescents globally.¹ The prevalence of off-label prescriptions has been shown to be between 3% and 95% in children in general in different settings worldwide, between 26% and 95% in newborns,² between 23% and 51% in outpatient children³ and between 9% and 79% in hospitalised children.³ In China, off-label drug use is common in paediatrics, one of the main reasons being the severe shortage of approved medicines for children. According to the China National Association of Industry and Commerce Pharmaceutical Chamber,⁴ there are over 6000 pharmaceutical companies in China, of which only about 30 are producing paediatric drugs. Until February 2021, there were a total of 157 790 domestic and imported medicines available in China, but only 3334 (2.11%) were dedicated to children, which is far from enough to meet the therapeutic demand.⁵ Although the Chinese government has made significant efforts in the previous 8 years to support the development of pharmaceuticals for children,⁶ the shortages in approved paediatric medicines persist.

Off-label drug use in China is defined as a drug not being used according to instructions approved by the National Medical Products Administration (NMPA), the authority responsible for drug licensing. Off-label drug use should be based on evidence, be beneficial to patients and be neither unethical, illegal, dangerous nor experimental.⁷ However, in clinical practice, much of the off-label drug use in paediatrics is not supported by adequate evidence and may thus be inappropriate and potentially even harmful to the patient.^{8 9} As a result, it is of critical importance to understand the current situation of non-evidence-based off-label drug prescriptions in paediatrics in China.^10–13

A number of studies focusing on off-label prescriptions in paediatrics have been conducted in China, but most of them are single-centre cross-sectional studies with small sample sizes^{2 14} or did not evaluate the evidence basis.^{3 15} The costs associated with off-label drug use in children are estimated high globally,^{16 17} but so far little is known about the situation in China.

This study aims to provide a comprehensive picture of the current situation with paediatric evidence-based and non-evidence-based off-label drug use in Gansu, China, as well as to identify prospective factors that could impact the management of paediatric off-label prescriptions.

Materials and methods

This study reviewed all prescriptions for children aged <18 years from 2 to 8 March and from 1 to 7 September 2020 in all secondary and tertiary public hospitals in Gansu Province, with the exception of hospitals of Tibetan medicine. A total of 196 medical institutions were included in this study, including 157 secondary hospitals and 39 tertiary hospitals. According to the Prescription Administrative Policy, a prescription is a medical document issued by a clinician for a patient in diagnostic or treatment activities, which is examined, dispensed and checked by a pharmacist, and serves as a voucher for the patient’s medication. Each prescription should contain up to five medicines.

In this multicentre cross-sectional analysis, the prescriptions were evaluated according to instructions approved by the NMPA. The following indicators were collected: whether the prescription was off-label, and if yes, the type of off-label use (please see the categorisation in the 'Determination of off-label use' section); the drugs used off-label; the cost of the drugs and selected characteristics potentially influencing the probability to prescribe off-label drugs. We followed the STROBE statement in reporting.¹⁸

Information extraction and data collection

We exported all paediatric (aged <18 years) outpatient prescriptions between 2 and 8 March and between 1 and 7 September 2020, from 196 secondary and tertiary hospitals in Gansu Province from the Hospital Information System.

Outpatient prescriptions were collected according to their date of issue. There was no follow-up on whether and when the patient received the medicine. Data were collected retrospectively in August 2021. Cleaning of the data was completed in October 2021, review of the prescriptions in April 2022, data analysis in June 2022 and article writing in August 2022.

We extracted the following information from each prescription: hospital name, visit identification number, patient’s sex, date of birth, date of visit, department visited, prescribing clinicians’s position/experience, disease or condition, and information about the drug (name, manufacturer, route of administration, frequency of administration, dosage, number of doses dispensed and total amount of the drugs prescribed).

Disease types were classified according to the International Classification of Diseases 11th edition (ICD-11).¹⁹ Patients were divided into four age groups²⁰: neonates (0–27 days), infants (28 days–23 months), children (2–11 years) and adolescents (12–17 years).

Inclusion and exclusion criteria for the evaluation of prescription

The following types of prescriptions were excluded: (1) Prescriptions of herbal medicines such as Astragalus, Angelica and others, which are usually complex and need to be determined according to the theories of Traditional Chinese Medicine; (2) Hospital-specific tailored preparations without instructions, approved by the NMPA, that are only allowed to be used in the hospital that developed them; (3) Healthcare products, that is, products that are not medicines used for other reasons than treating a disease; (4) Non-pharmaceuticals, such as disinfectants, gauzes or syringes and (5) Incorrect or incomplete prescriptions, with insufficient information to determine whether the use was off-label or not because of, for example, missing age, or diagnosis.

Determination of off-label use

Six experienced clinical pharmacists were divided into three groups. Each group was given a part of the extracted prescriptions for evaluation. In all groups, investigators independently evaluated the prescriptions and solved inconsistent judgments by discussion or consulting experts. To ensure sufficient consistency among the investigators, a pilot phase of 100 prescription assessments was conducted. The formal test began when the intraclass correlation coefficient of the evaluations among the six investigators was greater than 0.75.

First, we determined whether the drug was used off-label or not, referring to the instructions of the NMPA in the MCDEX database. The MCDEX database is authoritative and reliable and considered the best clinical medical information database system in China, similar to the ‘The NICE British National Formulary for Children’.

Second, we determined whether the off-label use of drugs was based on adequate evidence. The Benefit and Risk Assessment for Off-label use (BRAvO) decision-making framework is the first hands-on tool to be used in a clinical setting, and it is also widely used in inpatient children. However, as our study was a retrospective study of outpatient prescriptions, we did not have sufficient information to apply the BRAvO framework. Therefore, we determined adequate evidence as either having a Thomson grade IIb or above according to the Micromedex 2021 compendium (Micromedex database; online supplemental table 1),²¹ or being supported by clinical practice guidelines in the MCDEX database.²² The Micromedex database is an evidence-based, synthesised factual database that provides healthcare professionals with answers to specific clinical questions on pharmaceuticals. We used the Thomson grading system of Micromedex database to determine the adequacy of the evidence basis, including recommendations, evidence and efficacy rating.

Third, we determined the type of off-label drug use into the following seven categories according to NMPA’s instructions²³: (a) unapproved indication (the drug was used for a condition not mentioned in the drug’s instructions); (b) unapproved age (the drug was administered to children outside the age range specified in the instructions); (c) unapproved population group (the drug is approved for the relevant indication in adults but not in children); (d) unapproved route of administration; (e) unapproved dosage or frequency (dose and frequency inconsistent with the instructions); (f) contraindication (the drug is given with contraindication) and (g) the solvents used in drug preparation differed from the recommendations of the drug instructions.

Statistical analysis

Python V.3.6.5 was used to synthesise the data, SPSS V.23.0 to analyse and process it and Excel V.16.54 to present the results graphically. For quantitative data with a skewed distribution (eg, prescription costs), we reported the median and IQR and used the rank-sum test for comparisons. Categorical data (eg, off-label drug prescriptions) were presented as frequencies and percentages. The total costs of drugs between secondary and tertiary hospitals were compared using the χ² test.

We used univariable and multivariable logistic regression to investigate the role of potentially relevant factors on off-label prescriptions. Level of the hospital,²⁴ patient’s age group,²⁵ the position of the prescribing clinician²⁴ and whether the medicine was topical²⁶ were reported as influencing factors for off-label drug use in children and thus considered in the models. Then logistic regression analysis was used to investigate the role of the above-influencing factors after single-factor regression analysis. A forward stepwise likelihood ratio test was used to decide on the inclusion of variables with the rejection levels (α_in=0.05 and α_out=0.10) in the final multivariable model.

Results

Characteristics of the prescriptions

A total of 124 526 prescriptions (containing 255 793 medical prescription records of single drugs) were collected from 196 hospitals: 157 secondary hospitals with 74 859 prescriptions and 39 tertiary hospitals with 49 667 prescriptions (table 1; online supplemental figure 1). Of the 255 793 records, 46 323 records were excluded: 9039 records of patients aged 18 years or above, 575 records of non-pharmaceutical prescriptions, 909 records of hospital preparations, 21 525 records of Chinese patent medicine prescriptions, 148 records of other traditional Chinese medicine prescriptions and 14 127 records of inappropriate prescriptions. Finally, 104 029 prescriptions (containing 209 470 medical records of single drugs) were included, with 42 740 prescriptions from 39 tertiary hospitals and 61 289 prescriptions from 157 secondary hospitals.

Table 1.

Characteristics of the included prescriptions

Characteristics	Secondary hospitals (157 hospitals; N=61 289)	Tertiary hospitals (39 hospitals; N=42 740)	Total (196 hospitals; N=1 04 029)
Sex
Male	34 945 (57.0)	24 566 (57.5)	59 511
Female	26 344 (43.0)	18 174 (42.5)	44 518
Age
0–27 days	1712 (2.8)	626 (1.5)	2338
28 days–23 months	17 588 (28.7)	12 415 (29.0)	30 003
2–11 years	33 578 (54.8)	23 189 (54.3)	56 767
12–17 years	8411 (13.7)	6510 (15.2)	14 921
Prescriptions containing certain types of medicines
Prescriptions containing antibiotics	20 815 (34.0)	11 018 (25.8)	31 833
Prescriptions containing vitamins or trace elements	14 322 (23.4)	6230 (14.6)	20 552
Prescriptions containing steroids	4893 (8.0)	5014 (11.7)	9907
Costs of prescriptions (¥)	27.3 (9.3, 52.8)	46.5 (21.3, 96.3)	33.9 (14.8, 68.2)
Prescriptions containing antibiotics	32.5 (17.2, 58.0)	52.9 (25.0, 103.3)	38.0 (20.3, 71.5)
Prescriptions containing vitamins or trace elements	28.8 (13.7, 52.0)	59.2 (26.6, 107.7)	35.8 (17.2, 67.4)
Prescriptions containing steroids	40.5 (20.0, 74.3)	74.1 (36.3, 136.4)	57.3 (25.7, 104.4)

Costs of prescriptions are presented as median (IQR), other variables as absolute numbers and column percentages (n (%)).

Diagnosis of the off-label prescriptions

A total of 39 480 (38.0%) off-label prescriptions were identified among the 104 029 included prescriptions. The prevalence was similar in secondary (n=24 881, 40.6%) and tertiary (n=14 599, 34.2%) hospitals. The off-label prescriptions covered a total of 22 ICD-11 disease areas, of which the most frequent were diseases of the respiratory system (n=15 831, 40.1%), symptoms, signs or clinical findings, not elsewhere classified (n=6974, 17.6%), diseases of the skin (n=3079, 7.4%), diseases of the digestive system (n=2740, 6.4%) and diseases of the visual system (n=2674, 6.8%; figure 1).

Figure 1.

Diseases and conditions addressed in the included off-label prescriptions.

Types of off-label drug use

About a quarter (n=10 130, 25.6%) of the off-label drug prescriptions were supported by evidence (online supplemental table 2). The proportion of evidence-based off-label prescriptions was similar between secondary (n=6357, 25.5%) and tertiary (n=3773, 25.8%) hospitals.

The most common types of off-label drug use were unapproved indication (n=25 891, 65.6%) and unapproved dosage or frequency (n=16 622, 42.1%). 31% of the prescriptions with unapproved indications and 22.2% of the prescriptions with unapproved dosage or frequency were evidence-based (figure 2). Of note, a single prescription may be associated with more than one type of off-label use, with 27 835 (70.6%) fulfilling the criteria for at least two types and 4541 (29.5%) for at least three types simultaneously (online supplemental table 3).

Figure 2.

The types of off-label drug use in the included prescriptions.

Drugs used off-label

A total of 1171 drugs were addressed in the off-label prescriptions. Multienzyme tablets (n=1790), montelukast sodium chewable tablets (n=1317) and amoxicillin and clavulanate potassium granules (n=1141) were the most frequently prescribed off-label drugs (online supplemental table 4).

Economic burden of off-label drugs

The median (IQR) cost of the off-label drug was lower in secondary than tertiary hospitals 6.1 (1.1, 23.8) and 29.0 (9.2, 48.0) CNY, respectively; p<0.001). The estimated total daily cost of paediatric drugs prescribed off-label in Gansu Province was ¥43 241 in secondary hospitals and ¥62 876 in tertiary hospitals (table 2). The estimated daily cost of paediatric drugs used off-label without adequate evidence in Gansu Province was ¥81 941.

Table 2.

Costs of drugs prescribed off-label

Hospital level	Median costs (¥)			Total cost (¥/day)
	OLD	EOLD	NEOLD	OLD	EOLD	NEOLD
Secondary	6.1 (1.1, 23.8)	11.4 (0.8, 27.4)	6.0 (1.3, 22.8)	43 240.7	9626.2	33 614.6
Tertiary	29.0 (9.2, 48.0)	28.6 (6.6, 49.6)	29.0 (9.5, 47.7)	62 875.5	14 549.5	48 326.0
P value	<0.001	<0.001	<0.001	<0.001	<0.001	<0.001
Overall	14.5 (1.9, 32.3)	18.5 (1.1, 33.3)	13.0 (2.0, 31.7)	106 116.3	24 175.7	81 940.6

Total cost/day=the total amount of all off-label drugs given in the 196 hospitals divided by the total observation period (14 days).

EOLD, evidence-based off-label drugs; NEOLD, non-evidence-based off-label drugs; OLD, off-label drugs.

Analysis of influencing factors

According to the univariable analysis, hospital level, patient’s age, the prescribing clinician’s position and the medicine being topical were all independently associated with the prevalence of off-label drug use and thus included in the multivariable analysis.

Off-label prescriptions were less common in tertiary hospitals than in secondary hospitals (aOR=0.80, 95% CI 0.78 to 0.82; online supplemental table 5). Compared with the other age groups, adolescents had the highest prevalence of off-label drug use (aOR=1.04, 95% CI 1.02 to 1.06). Topical medicines were more frequently prescribed off-label than non-topical medicines (aOR=1.58, 95% CI 1.53 to 1.63). Senior-level clinicians prescribed drugs off-label more often than intermediate and junior clinicians (aOR=1.13, 95% CI 1.11 to 1.15 compared with junior clinicians).

Discussion

This cross-sectional analysis reviewed 104 029 prescriptions from 196 hospitals, with nearly a third of them including off-label drug use. Respiratory disorders were the most common conditions to be treated with off-label drugs. Around a quarter of off-label drug prescriptions in secondary hospitals and one-third in tertiary hospitals were evidence-based. Unapproved indication was the most common type of off-label drug use. We identified more than 1000 different medications being prescribed off-label, with multienzyme tablets being the most common. The total cost of non-evidence-based off-label prescriptions for children in Gansu, a province of China with 25 million inhabitants, was estimated to be about ¥82 000 every day. The majority of the prescribed medications were topical medicines. Senior clinicians tended to write off-label prescriptions more often than junior clinicians.

The prevalence estimates of off-label use from other studies in China are in line with our findings. A recent single-centre study with 322 000 paediatric prescriptions found that the proportion of paediatric prescriptions that included off-label use was 21.7% in 2019 and 18.4% in 2020.⁹ A multicentre study of paediatric off-label medicine prescriptions in general tertiary hospitals in Guiyang revealed a prevalence of 37.2%.²⁷ According to a comprehensive survey of hospitals in China, outpatient paediatric off-label drug prescription rates ranged from 22.7% to 51.2%.³ We found that 38.0% of outpatient paediatric prescriptions in Gansu Province contained off-label drug use, which is in line with the prevalence estimates found in the other studies.^{3 27} Two single-centre studies from outpatient clinics in China found that 81.4%²⁸ and 33.9%²⁹ of the prescriptions for children, respectively, were off-label.

Off-label drug use should always be based on evidence.^{8 30} The Law on Doctors of the People’s Republic of China, which came into force on 1 March 2022, emphasises the requirements for appropriate drug use and the necessity of adequate evidence basis for off-label use.³¹ Based on a review of relevant clinical practice guidelines and the Micromedex 2021 compendia, we found that only a quarter of the paediatric off-label prescriptions we identified were evidence-based. According to a study of the evidence basis for off-label drug use in a maternal and child health hospital in China, instructions approved by the U.S. Food and Drug Administration (FDA) accounted for 19.0%, international or national guidelines or formularies for 28.6%, and expert consensus for 16.7%.

The proportion of paediatric off-label prescriptions that are evidence-based and the percentage of medicines that are used off-label both vary substantially across studies.^{15 32–34} The most likely reasons for the variability is that most existing studies were carried out in a single centre with a small-sample size, thus not being representative of the population. Moreover, there is inconsistency in the definition of evidence-based drug use in paediatrics.^{15 32–34} As a result, China needs to establish an evidence-based framework for the management of off-label drug use in paediatrics.

According to our study, multienzyme tablets are particularly commonly used without evidence in paediatrics. An Estonian study on paediatric outpatient prescriptions also found multienzymes the most commonly prescribed off-label drug in infants and children.³⁵ Only small-sample studies^{36 37} and one clinical randomised controlled trial³⁸ have proven the effectiveness of multienzymes in treating dyspepsia and anorexia in children. The use of multienzymes in the treatment of upper respiratory tract infections is not supported by evidence. Furthermore, montelukast is commonly used without evidence support. Current evidence supports the off-label use of montelukast to treat respiratory disorders such as capillary bronchitis, and obstructive sleep apnoea syndrome. However, there is no evidence for the efficacy of montelukast to treat bronchitis, upper respiratory tract infections or cough (except if due to asthma).³⁹ Similarly, the FDA has listed montelukast as an off-label drug for cough in paediatrics (without asthma). The FDA also put a ‘black box’, that is, the highest level of safety-related warning, on montelukast because it can cause severe neuropsychiatric problems in children who take it to treat their allergic rhinitis.⁴⁰

The lack of topical formulations approved for paediatric use in China has resulted in widespread off-label use. A total of 2495 paediatric medications were reviewed in a study of paediatric pharmaceutical formulations in China. However, the proportion of formulations specifically developed for paediatric use was 7.6%, of which only 8.9% were for topical use.²⁶ Another study found that 54% of the instructions on topical medicines lacked information on paediatric use.⁴¹ A review of paediatric topical skin treatments found that 47.8% included labelling for paediatric use.⁴² This investigation indicated that many topical drug instructions are labelled with vague and potentially confusing information related to paediatric use, such as ‘Please consult the clinician or pharmacist’. The few topical medicines approved for paediatric use are also often accompanied by insufficient information, which increases the need for off-label use.

The findings of this study have the following implications for off-label drug use in paediatrics in China. First, there is a need for an evidence-based medical system for paediatric off-label use of drugs. Second, the management of off-label drug use in children, especially for respiratory diseases, needs improvement. Third, evidence-based recommendations for drugs commonly used off-label should be formulated to reduce unnecessary medication use and waste of resources. Finally, more formulations for children should be developed, and instructions of existing drugs for children need improvement.

Our study has several strengths. Since automated or semiautomated evaluation relying only on ICD codes of the main diagnosis may not be able to accurately characterize the true reason for the medication use (which could be also, for example, for treating certain symptoms), our team of six pharmacists manually evaluated all prescriptions to ensure that our findings reflect the off-label use of drugs in real-world clinical practice as accurately as possible. Our study covered all secondary and tertiary medical institutions in Gansu Province, which can be considered a representative sample of Northwest China in general. We also investigated the factors influencing paediatric off-label drug use and calculated the cost of inappropriate prescriptions.

This study, however, has also limitations. Drug instructions and drug-related guidelines were obtained from the MCDEX database and the evidence grade was from the Micromedex database. These databases are not frequently updated and some of the recent evidence may be missed. Despite the large sample size, our study was conducted in a single province and during a short time period and thus may not be fully representative of the overall situation in China. We also analysed the influencing factors of inappropriate prescribing, identifying hospital level, different age groups, clinician’s title and whether it is a topical preparation are all influential factors. However, we did not explore the reasons behind these associations further as they were not within the scope of this study. A dedicated analysis on the reasons behind these associations will be conducted in the future.

Conclusion

Our study of outpatient paediatric prescriptions from 196 hospitals in Gansu Province, China, found that off-label drug use is common. While about one-quarter of the prescriptions were based on evidence, the remaining three quarters may potentially include inappropriate medication use, with a total cost of about ¥81 000, corresponding to about US$11 200, per day in Gansu Province, a province of China with about 25 million inhabitants.⁴³ The development of drugs for children, especially drugs for neonates and topical medicines, should be promoted. The management of off-label drug use in paediatrics in China needs improvement.

Data availability statement

Data are available upon reasonable request. Not applicable.

Ethics statements

Patient consent for publication

Not applicable.

Ethics approval

The ethical approval for the study was given by the Medical Ethics Committee of Gansu Provincial People's Hospital, Gansu, China (approval number: 2021-285). All prescriptions were collected by the Gansu Provincial Health Committee from various medical institutions. We confirm that all methods were performed in accordance with the relevant guidelines and regulations. The study was performed in accordance with the Declaration of Helsinki.