To the best of our knowledge, this study was the first to investigate the extent of off-label and unlicensed drugs prescribed in Neonatal Intensive Care Units (NICUs) in Iran. Our study was conducted in two NICUs, one with neonates referred from other centers (outborn) and the other with neonates born in the same hospital (inborn).
The mean gestational age of the neonates in our study was 34 weeks, which is similar to the results reported previously in Germany, Ireland and France [3, 16, 17]. The rate of cesarean delivery of the preterm neonates was higher in our study than that of the term neonates. (P-value = 0.002), which is similar to the results of a previous study in Portugal . The higher prevalence of cesarean delivery in high risk preterm neonates is clearly to prevent severe vaginal delivery complications.
In this study, the mean hospitalization length for neonates in the NICU was 10.6 days, which was the same as in a study by Jana Lass et al. in Estonia (10 days) and less than in some other studies [3, 17,18,19]. The most common cause of hospitalization for neonates in this study was sepsis, while in two similar studies, the most common causes of the hospitalization were neonatal hyperbilirubinemia and respiratory distress [19, 20].
In the present study, as in some previous studies, according to the Anatomical Therapeutic Chemical (ATC) Classification System, systemic antibiotics were the most commonly prescribed drugs (n = 463, 44.1%) [3, 21].
In our study, out of the total 1049 prescriptions, 38.1% were off-label, which is similar to the results of studies in the Slovak Republic (43%) , Spain (41.4%)  and Finland (36%) . It is a lower percentage than was found in other studies conducted in France (59.5%) , Portugal (52.7%)  and Australia (47%) ; however, it is a higher percentage than was identified in a study performed in Switzerland (25%) .
The unlicensed prescriptions in this study were 1.9% of the total prescriptions, which was significantly lower than the prevalence identified in most of the previous studies [5, 22,23,24].
This variation in results could be due to the application of different methods in this study and different definitions of off-label and unlicensed drugs used in other studies.
This study found that 85% of neonates admitted to the NICU received at least one off-label or unlicensed drug, which was a similar result to that of a study done in a NICU in Australia (80%)  and close to the results of some similar studies conducted in recent years (69.7-100%) [5, 17, 18, 24]. Ampicillin and Gentamicin were the most commonly prescribed off-label drugs in the present study, which are the same drugs identified in a study in Spain . This finding was different from that of a similar study previously performed by Salehifar et al. in Iran with the purpose of determination of antibiotics consumption, in which the most and the least frequently used antibiotics were ceftriaxone and gentamicin, respectively .
The high prevalence of prescribing off-label drugs is not only due to the lack of clinical evidence, but also to the lack of complete information in neonatal references. We showed in this study that the prescription of off-label and unlicensed drugs was significantly higher in term neonates than in preterm neonates (46.9% vs. 32.6%); this finding is similar to those of two studies conducted in recent years [19, 21]. However, among preterm neonates, off-label dugs were given to a higher percentage of newborns with gestational age of less than 32 weeks (87.2%) than to neonates with gestational age of 32-36 weeks (84.5%). In general, similar percentages of term and preterm neonates were exposed to off-label or unlicensed drugs, which indicated that in the present study, the gestational age did not significantly affect whether the neonates received this type of drugs. However, with regard to drug exposure, gestational age had a significant effect, and preterm neonates received more drugs than term neonates.
In this study the most common off-label drug prescribed was related to dose, which was similar to the findings of a study from Portugal,  and different from the findings of studies from Australia, Spain and Norway, in which the most common off-label prescription of drugs were related to indication, age and route of administration [20, 23, 26]. Because of the lack of suitable oral forms of drugs for neonates, splitting tablets and dissolving them in sterile fluids was the most common form of unlicensed prescription in our study. Dissolved Phenobarbital was the most commonly prescribed unlicensed drug in this study.
In the outborn group of neonates at the NICU, given the winter season and the epidemic of respiratory diseases (bronchiolitis), many neonates older than 7 days presented with respiratory distress symptoms, and Albuterol inhalation was repeatedly administered, which were prescribed off-label in 100% of cases, both in terms of dose and dosing interval. However, it seems that its administration was unnecessary for many neonates.
The frequent and long-term administration of antibiotics was found in both NICUs, which is unsurprising, because the most common diagnoses were sepsis, RDS and TTN. This use of antibiotics requires review and limitation according to in-hospital or national guidelines. The use of Bactec to perform cultures and to obtain their results in the shortest possible time should be considered, as well as the discontinuation of antibiotics in the event of negative cultures and the absence of risk factors.
This study showed that the prevalence of prescription of off-label and unlicensed drugs in our country is similar to that of reported in many other countries. The special conditions of neonates, especially preterm neonates, may be the main cause of these kind of prescriptions. To avoid this issue and to support the appropriate drug therapy in a high-risk group such as newborns, the first step is to provide relevant information about the risks and unwanted side-effects of these drugs to the physicians who are treating neonates. The second step is to carry out empirical studies and multiple clinical trials to develop low-risk medicines for neonates, such as the production of suitable oral formulations for neonates (oral solutions instead of tablets) by pharmaceutical companies to reduce the prescription of unlicensed drugs. We also believe that a neonatal formulary containing all the information needed on the clinical use of neonatal drugs in Iran, (as present in some European countries like BNFc (British National Formulary for Children) in the UK), is necessary. The presence of a clinical pharmacist in the NICU of the hospital is necessary to confirm the low risk or the prescriptive use of the prescribed medication.
The limitation of the present study is that it was retrospective, which led to the withdrawal of a number of files of neonates admitted to the NICUs. In addition, since this study was only conducted in one province, the findings about the prescription of off-label and/or unlicensed drugs cannot be generalized to the whole country and further studies are needed. As the final point, because of the lack of information about the immediate and long-term complications of off-label and unlicensed medications, there is no reliable evidence available about possible harmful side effects of these drugs.