Risk management
Measures to minimise foetal exposure to potential teratogens investigated in this review were based on components of Risk Evaluation and Mitigation Strategies (REMS) with Elements to Assure Safe Use (ETASU). These measures were: teratogenic counselling, contraceptive counselling, pregnancy testing before or at start of treatment, pregnancy testing while on treatment, use of contraception before or on starting treatment, and use of contraception during treatment. Since 2007, implementation of REMS with ETASU have been required by the FDA for medications with serious safety issues like teratogenic medications to ensure that the benefits of a medication outweigh the risks to patients [83].
Isotretinoin was the most commonly prescribed teratogenic medication covered by the studies included in this review (n= 16; 29.1%) [22, 30, 31, 33, 34, 47, 53, 63, 64, 66, 67, 71, 73-75, 77]. This may be because of two reasons. Firstly, isotretinoin is a relatively old medication that has been in the market since 1982-1983, and has been prescribed under a pregnancy prevention programme since 1988 [84]. Secondly, it is one of the most cost-effective acne treatments used by patients from different age groups including women of child bearing age [4, 77, 85].
By contrast, it was observed that there were fewer publications on medications prescribed under more recent risk management programmes such as thalidomide, lenalidomide, and valproic acid [18, 21, 41, 45, 59, 61, 70], indicating a need for further investigation of the safety of these medications in terms of adherence to risk management measures for such medications. Good practice guidance suggests that ensuring safe use of medications can have a number of positive effects on treatment outcomes. For teratogenic medications in particular, this includes reducing the incidents of medication- induced foetal harm and empowering patients to make the most of their treatment [11].
Results of teratogenic risk management implementation showed a wide variation among studies. Some studies reported surprisingly low rates of implementation. For example, only 9.5% of women of child bearing age using valproate received teratogenic counselling in the study by Mulryan et al. [70], and 6.1% of women using medications of class D or X received contraceptive counselling in the study by Schwarz et al. [17]. Additionally, rates of pregnancy testing before starting treatment with valproate or thalidomide were as low as 0% in two studies [59, 70], and pregnancy testing during treatment with acitretin was 12.7% in the study by Raguideau et al. [72]. Low rates of contraceptive use were also reported. Uuskula et al. reported that 15.7% of women of child bearing age on isotretinoin treatment in their study used a contraceptive before starting treatment [22], and Tsur et al. reported that only 1.7% of women in their study group used contraception during treatment with isotretinoin [75].
The wide variation in the results of implementing risk management measures can be discussed in the light of several factors. One factor could be the data sources used by the different studies. Some studies relied on medical records as their source of data [17, 19, 21, 22, 31, 32, 34, 36, 37, 40, 41, 43, 44, 46, 49, 54, 57, 58, 60, 61, 65-67, 70, 72, 78], while others used patients surveys [18, 20, 30, 33, 35, 39, 45, 47, 48, 52, 53, 55, 56, 62, 64, 69, 71, 74-77], a combination of medical records and patient surveys [38, 50, 63, 68, 73], or other sources (patient logs [42], reproductive life plans [51], and physician surveys [59]). Having patients as the only source of information can lead to several forms of self-reporting bias [86]. Recall bias can lead to an erroneous estimation of risk management variables if the patient’s recall of information is inaccurate [87]. Another form of self-reporting bias associated with the disclosure of sensitive data is the social desirability bias [86]. Social desirability bias might have led to an overestimated adherence to risk management and pregnancy prevention measures [88]. On the other hand, if data were extracted from the medical records, several issues like incomplete records, non-captured data, and low quality data might have an effect on the research outcomes [89, 90]. Therefore, it is important to bear in mind the possible types of bias associated with each source of data.
Another well recognised variable leading to variations in the implementation of risk management for the different teratogenic medications is the availability of risk management programmes. For certain medications like thalidomide, linaledomide, and isotretinoin, detailed risk management programmes that aim to prevent foetal exposure to the drug are in place [4, 8]. However, for other teratogenic medications, managing their risk is limited to the use of product labelling and patient information leaflets rather than rigorous monitoring [91]. The effectiveness of drug labelling as a risk management tool has been a matter of debate as research suggests a lack of effect on physicians’ prescribing behaviours or patients’ understanding of instructions [91].
Results of the current review can be considered as a compliance assessment of teratogenic risk management (whether through existing risk management programmes or through labelling recommendations) [91]. Based on the findings of this review, safety of the utilisation of teratogenic medications is sub-optimal, and entails a risk of foetal exposure to the harmful effect of potential teratogens. Consequently, it is recommended that the implementation of the existing teratogenic risk management programmes be monitored more carefully, and the criteria for the optimal management of teratogenic risk for potential teratogens be reviewed and revised based on the available evidence.
Exploring the implementation of risk management for teratogenic medications can help to develop interventions designed to minimise foetal exposure to cytotoxic effects, and thus future research utilising multiple data sources is needed. Drawing on the strengths of data extracted from medical records and patient reported data, mixed methods research that utilises quantitative and qualitative methods could yield more rigorous results than research utilising quantitative or qualitative methods alone [92, 93].
Consequently, results of this review raise two important issues. First, the review uncovers deficiencies in the implementation of risk management of teratogenic medications which constitutes a serious public health concern that needs further investigation. Second, it highlights a potential need to reinforce policies and regulations that aim to reduce foetal exposure to the cytotoxic effects of teratogenic medications.