Obstetric brachial plexus injuries (OBPIs): health-related quality of life in affected adults and parents

Obstetric brachial plexus injuries (OBPIs) are rare [1, 2] but can have significant implications for those affected [3], their caregivers [4] and the wider health system [5]. These injuries – sometimes referred to as Erb’s palsies – can occur as a result of the mismanagement of an obstetric emergency: shoulder dystocia; where the baby’s head is delivered but its anterior shoulder is impacted against the mother’s pelvic bone [6]. Inappropriate management of this emergency can lead to damage to the brachial plexus, a collection of nerves on either side of the neck which are responsible for sensation and movement of the arm [7]. Symptoms can range from restricted movement in the shoulder, elbow and wrist, to complete paralysis of the arm [1, 3]. They can also be transient or permanent (lasting for longer than 12 months after birth) [1, 8]. In England (2000 to 2010), litigation costs associated with shoulder dystocia and OBPIs amounted to over £103 million [5]. This is money that could otherwise have been spent on frontline patient care.

These potentially life-changing injuries warrant investigation into how they affect the quality of life of both sufferers and their caregivers using formal quality of life utility scores. In this context, utility is used to describe the preferences of individuals for a particular set of health outcomes and is way of valuing different health characteristics on a single scale [9, 10]. To date, the only quality of life utility scores related to OBPIs have been assigned at a distance, by healthcare professionals from outside the specialties involved in OBPIs [11]. There has been no direct assessment of the impact of OBPIs on individuals and although there is a growing body of research investigating the effects of OBPIs on families and caregivers [4, 12, 13], it has never been expressed using utility scores.

We used a generic, preference-based health status instrument (EQ-5D-5 L) to determine the quality of life utility scores of affected adults and parents of children with permanent OBPIs. The utility scores of these cohorts were compared with population norms. We also investigated whether certain socio-demographic or clinical factors were associated with the utility value.

Both adults with OBPIs and parents of children with permanent OBPIs had worse utility scores, and therefore quality of life, compared to the general English population. The mean utility score for affected adults was 0.56 (SD 0.28) and the mean utility score for the parents was 0.80 (SD 0.19). We also identified characteristics to consider when dealing with utility scores in populations of adults with OBPIs or parents of children with permanent OBPIs. For affected adults: previous OBPI surgery, employment status, disability benefits status and marital status appeared to be associated with the utility score. For parents: employment status and the presence of one or more parental medical condition appeared to be associated with the utility score.

Only one study has previously presented utility scores associated with OBPIs [11]. Culligan et al. assigned utility scores to different combinations of maternal-neonatal outcomes, some of which included permanent OBPIs [11]. A utility score of 0.60 was assigned by this group of clinicians to mild-moderate OBPIs, and it is similar to the affected adult utility score calculated in this study (7% difference). However, the utility score derived by this study is a more reliable and accurate measure of the quality of life associated with OBPIs. Firstly, the utility score was determined using a validated generic health instrument and involved adults with the injury. Secondly, the affected adult utility score in this study has a clear perspective. As the utility scores assigned by Culligan et al. are for combinations of maternal-neonatal outcomes, it is unclear whether the quality of life was being assessed from the viewpoint of the mother, the baby, or both [11].

The EQ-5D-5 L questionnaire was an appropriate instrument to use to calculate the utility scores in these cohorts. There are currently no OBPI-specific instruments and despite being a generic questionnaire, the domains covered by the EQ-5D-5 L, particularly self-care, usual activities, pain/discomfort and anxiety/depression were relevant to OBPIs. The best way of measuring and integrating the ‘spillover’ effects (the impact on caregivers and family members) of healthcare interventions into cost-effectiveness analyses is still debated [25] and generic health status instruments may not fully capture the ‘spillover’ effects, particularly the potential positive aspects of caring [25]. Despite this, the EQ-5D questionnaire has been able to detect changes in health-related quality of life in caregivers of children with major congenital anomalies [26] and it appears to be useful in measuring family members’ quality of life in economic evaluations [27]. The EQ-5D-5 L questionnaire is also the preferred health status instrument for NICE [28]. This means that the quality adjusted life years (QALYs) calculated from these utility scores would be comparable with the QALYs in other health technology assessments reviewed by NICE.

We mapped the EQ-5D-5 L responses to an EQ-5D-3 L valuation set as recommended by NICE using the eq5dmap command in Stata [19]. The command is a mapping process based on a system of ordinal regressions and does not assume that the domains of the EQ-5D instrument are statistically independent [19, 29]. We used the EuroQoL Group dataset [30, 31] as the reference for the mapping process. This dataset consists of a younger population and covers a broader range of diseases compared to the alternative reference dataset available (National Data Bank for Rheumatic Diseases) [32]. Furthermore, the EuroQoL Group data was generated from 6 European countries, including England, making it the more relevant reference dataset to use [30, 31]. The eq5dmap command also adjusts for age and gender [19].

The inability to stratify utility scores by the severity of the condition is a limitation of this study. Only a small proportion of affected adults (28%) and parents (41%) knew the Narakas classification of the injuries, even though it is widely used clinically [33]. As a result, a crude proxy variable was needed and previous OBPI surgery was a reasonable choice. There is some evidence that having lower Narakas scores (I-II) reduces the likelihood of permanent OBPIs, and hence the need for surgery [34]. Similarly, the presence of Horner’s syndrome, indicating Narakas score IV, increases the likelihood of permanent OBPIs and therefore operative management [34]. Previous OBPI surgery was associated with the affected adult utility score but not with the parent utility score in this study. In comparison, one previous study has demonstrated that OBPI surgery had an impact on the child’s family [12].

We did not attempt to measure the quality of life utility scores of the children with OBPIs. Young children may not have the cognitive capacity to comprehend or answer health related questions, and there are few preference-based, health status instruments derived from child populations [35, 36]. A small study (n = 18) has assessed the quality of life of adolescents (aged 10 to 17 years old) with permanent OBPIs using the Child Health Questionnaire and the Paediatric Outcomes Data Collection Instrument [37]. Whilst the questionnaire and instrument quantify health-related quality of life, the scores cannot be used to calculate QALYs as they are not preference-based. There is a youth version of the EQ-5D questionnaire (EQ-5D-Y) but there are no valuation sets for this at present. Using the EQ-5D adult values for EQ-5D-Y would misrepresent a child’s health status [38]. As a result, the EQ-5D-Y cannot generate reliable child QALYs for use in economic evaluations.

OBPIs are rare and as such, participants were difficult to recruit. Collaborating with the Erb’s Palsy Group allowed access to their membership, the majority of which would have been eligible to participate in the study. Despite attempts to maximise publicity and recruitment, the response rates were low. One reason for this may be because of the perceived lack of benefit in taking part in the study. The small sample sizes of affected adults and parents of children with permanent OBPIs limited the extent of our statistical investigations. As a result, our study had limited statistical power and we could not confidently detect some true associations. Moreover, our analyses were sensitive to the changes in the sample sizes (see additional analyses) and some associations were no longer evident in those additional analyses conducted on marginally larger samples. For example, previous OBPI surgery went from a positive association with the affected adult utility score to no association. This clinical factor was chosen as a proxy for the severity of OBPIs, but it could be possible that surgery improves function to the extent that it results in better quality of life. Further analyses with larger samples are needed to check whether there are indeed any associations between the factors identified in this study and the utility score in these cohorts. Interestingly, mothers of children with permanent OBPIs had a lower mean utility score (0.79) compared to fathers (0.89). One possible explanation for this could be that mothers might have a greater share of the caring responsibility. It is worth noting, however, that this study had a very small cohort of fathers (8% of the total sample) and an even smaller number of single fathers (1% of the total sample) to test this theory. Further research is needed to investigate whether there is a true difference in the quality of life between mothers and fathers of children with permanent OBPIs.

Due to the cross-sectional nature of the study, there may have been selection biases so the samples in this study may not be fully representative of all adults in the UK with OBPIs or parents of children with permanent OBPIs. Furthermore, as this study was based on questionnaires, the responses may also have been affected by recall biases.

Despite the small sample sizes, this is the first study to directly and robustly assess the quality of life of both adults with OBPIs and parents of children with permanent OBPIs using a validated health-status instrument. The findings from this study represent the best available evidence and provide tailored utility scores that could now be used as part of cost-utility analyses of interventions related to improving the care, health and wellbeing of adults affected by OBPIs and parents of children with permanent OBPIs.

Adults with OBPIs and parents of children with permanent OBPIs reported worse utility scores, and therefore quality of life, compared to the English general population. Previous OBPI surgery, employment status, disability benefits status and marital status appeared to influence the affected adult utility scores. Parental employment status and the presence of one or more parental medical condition appeared to have an association with the parent utility scores. The utility scores in this study can be used in future economic evaluations related to OBPIs.