For paediatric medicine to advance, research must be conducted specifically with children. Concern about poor recruitment has led to debate about payments to child research participants. Although concerns about undue influence by such ‘compensation’ have been expressed, it is useful to determine whether children can relate the time and inconvenience associated with participation to the value of payment offered. This study explores children's ability to determine fair remuneration for research participation, and reviews payments to children participating in research. Forty children were interviewed before outpatient visits at two London Hospitals: Great Ormond Street Children's Hospital and the Whittington Hospital District General Hospital. Children were asked to value their involvement in two hypothetical research scenarios – the first an ‘additional blood sample’, the second also involving daily oral oil capsules taken for a fortnight before further venesection. Background knowledge about familiarity with money, and experience with hospitalisation was assessed. The mean valuation of involvement in the second scenario (£13.18) was higher than in the first (£2.84) (p<0.001). This higher valuation persisted when children were categorised into groups ‘aged 12+’ and ‘below 12’. Those undergoing a blood test on the day placed a higher valuation on participation in the second scenario (£10.43, £21.67, p=0.044). These children aged 8–16 demonstrated the capacity to discern a fair valuation for participation in medical research. The monetary sums are influenced by the time and inconvenience involved in the research, and by the extent of recent experience with hospital procedures. The authors review current ethical thinking regarding payments to child research participants and suggest that a fair wage model might be an ethically acceptable way to increase participation of children in research.
- research ethics
- research on special populations
- policy guidelines/inst. review boards/review committees
Statistics from Altmetric.com
- research ethics
- research on special populations
- policy guidelines/inst. review boards/review committees
Historically, the lack of medical research involving children has meant paediatric practice is often built on extrapolations from adult studies. This is suboptimal as adult physiology, for example the rate and mechanisms of drug clearance, may differ greatly from that of children and interventions suitable for adults may not meet the psychological or physical needs of some children (eg, swallowing a tablet).1 Therefore relying on adult data may be insufficient or even unethical.2 While the recent increase in research with children, due to better funding of studies and of organisations such as the UK Medicines for Children Research Networks (MCRN), is encouraging, some are concerned about possible erosions of the protection of the child participant.3
The short project we describe was designed to obtain quantitative data from children and to estimate their understanding of the material worth of money as applied to two ‘pretend’ research scenarios. We have also sought to review the current ethical thinking regarding payments to child research participants.
The major ethical challenge in undertaking paediatric research is the participant's lack of capacity to give fully informed consent. Therefore, there is increased reliance on other protections for the child subject: all human subject research is reviewed by an independent National Research Ethics Service (NRES) Research Ethics Committee; paediatric researchers – as others – must undergo good clinical practice (GCP) training; and researchers are increasingly part of umbrella organisations such as MCRN. Generally those undertaking research with children seek consent from parents or adults with parental responsibility, a step mandated for incompetent children and all children involved in Clinical Trials of Investigative Medicinal Products (CTIMPS). Parents or adults with parental responsibility give fully informed consent as they are presumed to act in the child's best interest, although whether such decisions are automatically in a child's best interests or even what these interests are is debated.4
Children under the age of 16, if deemed competent, can independently consent to medical procedures held to be in their ‘best interests’. This ‘Gillick-Competency’ suggests, under common law, that children who have achieved certain levels of emotional and cognitive maturity and who are able to meet a number of criteria concerning their understanding of choices in the medical situation, intelligence, decision-making, and the ability to express reasonable views clearly may give consent.5 This standard is often applied to medical research that does not explicitly fall under clinical trials regulations.
Inevitably any determination of competency relies on a value judgement by the clinician,6 with capacity to make decisions depending on individual age, rationality and intelligence.7–9 Children often reach the same treatment decisions as adults, and although age is important, the capacity to make autonomous decisions is dependent on other factors such as education and, above all, experience with medical environments and procedures.10
Even with parents' consent, research guidance suggests children should be given age/developmentally appropriate information to enable them to understand the study and have an opinion about participation, which with increasing maturity ought to become determinative. Researchers must ensure children agreeing to participate have understood the information given to them and are deciding to participate without duress.11 Terms such as assent and child agreement are now common but remain poorly defined and controversial.12
For the scientific community, while participant protection is necessary, it must be balanced by potential detrimental effects on recruitment of scarce child research participants and consequent risk of perpetuation of poor evidence for child health.
At a recent NRES meeting on optimising recruitment to paediatric research,13 attendees debated direct payment to children, which led to this study. We investigated whether the offer of payment, common in adult studies,14 could increase paediatric research recruitment by helping children understand the value of participating in research. Since such payment may reimburse research participants for financial costs, time and effort, it is important that the rate of pay is fair and ethical.
Reimbursement payments compensate participants for direct financial expenses of involvement in the study. Such costs will only be based on the actual expenses incurred by the participant as a result of participation, and are a ‘revenue neutral’ form of payment,18 with no financial gain.
Compensation payments similarly reimburse direct costs of participation, but also reward participants for the time cost and effort of involvement in the study. The level of compensation offered depends on factors such as frequency of clinic visits, risk associated with involvement19 and length of hospital stay.
Appreciation payments are small tokens given after study completion to reward participants for their efforts. These may commonly take the form of gifts or vouchers, but money may also sometimes be offered.
Incentive payments are designed to directly encourage enrolment20 through the promise of financial gain above the direct and indirect expense of participation.
Conflicting guidance on research payments to children has been issued: The Royal College of Paediatrics and Child Health states that no financial inducements should be offered for fear of unduly influencing the child and parents.21 In contrast, the National Children's Bureau Guidelines suggest that studies offering some form of inducement should be considered,22 but not incentive payments as financial gain is always greater than direct costs and inconvenience.23 Recent legislation has emphasised the former viewpoint, the European parliament prohibiting all incentives or overt financial inducements in paediatric research.24
Revenue neutral reimbursement payments, with compensation for direct costs incurred, provide no possibility of financial gain unduly influencing decisions to participate.18 However, other than removing immediate financial barriers reimbursement offers no particular encouragement to participate. In turn, appreciation payments might act as incentives if parents or children are informed before enrolment.18 ,23 Compensation payments, however, may well exceed the cumulative direct and indirect costs of participation and thus unduly influence enrolment.23 Personal values and level of experience with hospitals and procedures surely also play a role in whether a child views a compensation payment as an incentive or not.18
The wage-payment model was devised in an attempt to encourage recruitment while minimising undue influence on the participant, who would be paid a sum of money in the form of an hourly wage. This form of payment could be considered a compromise between inducement and compensation, as it enables the participant to relate the payment offered to the time and effort involved in research, and therefore requires a concrete concept of the value of involvement. It has been suggested that in order to minimise undue influence payment should be calculated to correspond to the wage of unskilled workers.25 If such an approach is reasonable for adults, could it be considered in order to facilitate recruitment to paediatric studies—perhaps using the national minimum wage for the age group concerned? The success of any such model depends upon children's ability to understand the meaning of a wage. The concept of money forms by the age of 8–9 years, with some variation in the progression of understanding,26 and therefore evaluation of children's comprehension of a wage payment model of compensation is necessary.
The wage payment model has been evaluated in a qualitative study of 42 children and adolescents of various medical backgrounds.27 Four hypothetical research scenarios, with various amounts of inconvenience, effort and invasiveness, were devised. Interview transcripts containing the children's attitudes towards participation, and their own notion of what a fair wage payment would be for participation, were analysed. As with the criteria required for a child to be deemed Gillick competent,6 some key concepts must be understood by potential participants if the wage payment is to be considered appropriate. The most important concepts26 ,27 were the capacity of the child to demonstrate:
‘An understanding of how one earns or obtains money.’
‘The ability to correlate a given amount of money with its material worth.’
‘The capacity to discern what is fair payment for performing a given task for a certain amount of time.’
Children below the age of 9 years failed to meet some or all of these criteria, while most children above that age had an accurate concept of the material worth of money, and the ability to demonstrate clear rationale in asking for reasonable levels of wage payment in relation to the inconvenience of their participation in research.25 This would suggest that a wage payment model is suitable in children of our chosen age range.
Research ethics application
The South East London 3 proportionate review subcommittee of NRES gave ethical approval.
Identification, recruitment and consent
Children aged 8–16 (inclusive) were recruited from two London hospitals: the Whittington Hospital in Archway and Great Ormond Street Children's Hospital (GOSH). To find and recruit children for the study, lists were accessed 1 week in advance of a given clinic with the assistance of clinical staff at each site. A letter describing the interview was sent to parents/guardians and children in the correct age range several days before the clinic, enclosing two information sheets: one for the parents, one for the child (separate versions for 8–12 year olds and for 12+ year olds).
In total, we wrote to 117 families. Of these, 53 were found on the day of their clinic and approached to participate in the study. Children were considered unsuitable if overtly distressed or their cognitive status precluded completion of the questionnaires. Of the 50 considered suitable, 10 declined to participate; some families were unwilling to spend longer in hospital than necessary, or said the child was already distressed by a blood test. Those willing to participate were required to give consent: one consent form for a parent and either an assent form (for children under 12 years) or a consent form (for older children). Duplicates of these forms were handed to the parents. The interview began at a convenient time after the initial part of the outpatient appointment, which often included a blood sample and routine measurements.
A scripted interview was administered verbally by the same researcher in all cases, in the presence of the parents. The child was allowed to drop out after the introductory stage in which age and sex were recorded, and at the end of the first scenario. Answers were recorded on the script in the sight of the participants. The identity of the participant was not linked to the names given in the consent stage.
The interview (online appendix 1) included questions concerning the child's familiarity with money, experience with hospitalisation, whether they had given a blood test, and their opinion and valuation of two hypothetical research scenarios. The first scenario involved giving an additional small amount of blood with a blood test in the routine clinic. It was emphasised that this was ‘pretend’. The second scenario was designed to be rather more invasive, and involved giving this additional small amount of blood, and then sending the child home for 2 weeks to take an oil tablet daily, before returning to the hospital to have a ‘research’ blood test. In each case payments in the form of vouchers were mentioned and the child was invited to pick the ‘right’ value.
A statistical analysis software package (SPSS version 18) was used for data analysis using non-parametric tests, including the Mann–Whitney U test, Wilcoxon signed rank test and Kruskal–Wallis test due to the small sample size and distribution of the data.
Results and analysis
Descriptive statistics of sample
The final sample of 40 children comprised 22 girls and 18 boys. Twenty-four (60%) were recruited from the GOSH rheumatology clinic, three from the GOSH transplant clinic and the remaining 13 (32.5%) from the Whittington phlebotomy clinic.
Rheumatology and phlebotomy clinics displayed a fairly even spread of ages, with the phlebotomy clinic having the lowest mean age of recruitment (12.33) and the transplant clinic the highest (16.39), albeit with only three participants.
Although it was anticipated that the phlebotomy clinic would provide less ‘experienced’ patients compared with perhaps more complex patients attending specialist clinics at GOSH, who would be more familiar with hospitalisation/medical procedures, there were no striking differences in distribution of ‘experience’ (recorded by number and nature of past hospital visits) between the rheumatology and phlebotomy clinics (figure 1). The three transplant clinic patients showed a higher degree of hospital experience.
We found that few of these children received money as a function of work, either through household chores or through a paid job. Twelve participants reported no regular encounters with money. Most of the participants received occasional small amounts for holidays/birthdays, or received weekly/monthly pocket money (figure 2).
Willingness to participate in hypothetical scenarios
The study intended to determine what value children would place on their participation in hypothetical research scenarios. Initially, however, their willingness to participate regardless of a wage payment was tested. Willingness to participate in both of the hypothetical research scenarios was high, with 82.5% of recruited patients saying they would be ‘fine’ with participating in the first hypothetical scenario, and 70% saying they would be ‘fine’ with participating in the second hypothetical scenario.
Unwillingness (‘definitely not’ response) to participate in the second scenario (20%) was twice as high as in the first (10%). There appeared to be small discrepancies between boys and girls when it came to willingness to participate in either of the hypothetical scenarios, with boys less likely than girls to be ‘fine’ with participation, and more likely to respond that they were ‘definitely not’ or ‘possibly’ interested compared with girls. This difference was not significant at the 95% level, however.
We found a trend with age when willingness to participate was established before wage payment was mentioned (figure 3). The mean age of patients definitely willing (‘fine’) to participate in scenarios 1 and 2 (13.45 and 13.25) was higher than the mean age of patients definitely unwilling to participate in the scenarios 1 (11.25 and 11.97 respectively). This difference was not significant at the 95% level, however. The second scenario also attracted a higher number of responses in the categories of ‘possibly’ (4) and ‘definitely not’ (8) than for scenario 1 (3 and 4 respectively).
We found no statistical evidence for a link between extent of hospital experience, whether the patient had given a blood test on the day and willingness to participate in either hypothetical research scenario.
Value placed on participation in hypothetical scenarios
The primary research question concerned the value that children placed on their participation in two hypothetical research scenarios. The first scenario specified less time and effort than the second. The mean valuation of the first scenario, including children who specified gratis involvement, was found to be £2.84, and for the second £13.18 (see table 1). The difference in these two valuations was statistically significant (Wilcoxon signed ranks test, p<0.001). When analysis of responses was performed again on the children aged 12+ and again on the younger group the mean valuation increase between the two hypothetical scenarios for each of these groups was still found to be significant (p=0.001 and p=0.042 respectively), although four of the younger children dropped out before giving a financial estimate for scenario 2.
On the assumption that the first research scenario offers a one-off payment, and the second research scenario would require, on average, 8 h and 20 min of each participant's time (6 h for the return visit to hospital and 5 min taking a tablet twice a day for 14 days), an hourly wage can be determined from the mean valuations. For the second scenario this is estimated as £1.58 an hour. When data were split once more into groups, the older children's mean valuation of scenario 2 amounted to an hourly wage of £1.49 and the younger group's mean valuation amounted to a higher hourly wage of £1.80. When children who said they would participate in scenario 2 gratis were removed from the analysis, the hourly wage derived from the mean valuation of scenario two is £2.17 an hour, with a mean hourly wage of £2.25 and £2.03 for older and younger children respectively.
On the day of the interviews each child may have had a blood test. For scenario 1, the valuation placed on participation by children who had not received a blood test (n=31, μ=£2.39) was not significantly different from the valuation placed by children who had received a blood test (n=9, μ=£5.00) (Mann–Whitney U test, p=0.101). However, the valuations between these two groups for scenario 2 (mean valuation of £10.43 for children who had not received a blood test and £21.67 for children who had) showed a statistically significant difference (p=0.044).
Relative hospital ‘experience’ was also investigated for its effect on valuation of the hypothetical research scenarios. We found no clear link between experience and the valuation of scenario 1, but there appeared to be a relationship between increasing degree of hospital experience and a higher valuation of scenario 2. However, upon conducting a Kruskal–Wallis test (chosen due to the non-parametric nature of the data and more than three groups for analysis) this trend was not found to be statistically significant (p=0.350).
While this study surveyed a relatively small number of children, and any quantitative conclusions are of course limited by this, there is little existing research regarding payments to children for involvement in medical research.
By excluding children below the age of 8 years, we ensured that all subjects had the ability to understand the concept of a wage.25 ,26 We intended to focus on a quantitative value of a fair wage payment and on how this changed with increased inconvenience, invasiveness and time cost of participation. It was hypothesised that a child with a greater level of experience with procedures and time spent in hospital might view participation in the hypothetical scenarios as less trouble than a relatively ‘inexperienced’ child, and would reflect this in the valuation. However, a child who had experienced a relatively traumatic procedure or event on the day of the study might well place a higher value of wage payment on participation in the hypothetical scenarios. Experience and familiarity with money could themselves be considered a function of age, along with increased maturity, so age was also included in the analysis. Age is strongly related to ‘life experience’, which may be more significant in a child's attitude towards participation in medical research than the ‘experience of hospitalisation’ recorded in this study. Finally, the sex of the participant was analysed.
We were careful to ensure at all stages that the hypothetical nature of the scenarios was understood: with younger participants words such as ‘pretend’ and ‘imagine’ assisted in their understanding, with parents free to explain throughout in terms that their child would easily understand. A single interviewer was used throughout to minimise bias.
Willingness to participate in each scenario was assessed before mention of compensation; in general, older children were more likely to express willingness than younger children. The question of payment generated a mean valuation of £2.84 for the first scenario compared with £13.18 for the second (p<0.001). It seems evident that the extra time and inconvenience in the second scenario influenced children to place a higher valuation on their participation than in the first. The actual valuation of scenario 2 was higher in the younger group, who may have been more daunted by the extra effort described.
We calculated a notional wage from the mean valuation of scenario 2, assuming an additional 8 h and 20 min of ‘work’. The mean hourly wage for participation in scenario 2 was £1.58 an hour. However, when the children who offered their participation gratis were removed from the analysis the mean notional wage increased to £2.17 an hour. While there is no defined minimum wage below the age of 16 in the UK, the minimum wage28 for 16–17 year olds stood at £3.64 an hour at the time of the study. An hourly rate of £2.17 seems, therefore, a reasonable wage payment for children aged 13–14, the mean age in the sample. This suggests that children can determine a fair valuation for their participation on the basis of a wage payment model in which compensation is directly correlated to the time and inconvenience of their participation.
We found that a higher valuation of scenario 2 was more closely related to having had a blood test shortly before the interview (significant at the 95% level, p=0.044) than the trend for older children to mention a higher value.
In summary, this study suggests that children aged 8–16 years have the capacity to discern a fair wage payment for their involvement in medical research. We calculated a mean hourly wage of £2.17 based on the time and inconvenience involved for participation in scenario 2. This figure is close to the minimum wage for children in the age range found in this study, and so demonstrates the potential suitability of a wage payment model in allowing children to determine the value of their participation in research, without incentivising such participation. The influence of hospital experience on valuation should be explored further to determine whether the short-term effect, as shown by the response to recent blood tests or lengthy experience of inpatient and outpatient procedures, affects children's rationalisation of the value of participating in research.
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Files in this Data Supplement:
- Data supplement 1 - Online appendix
Competing interests None.
Ethics approval South East London 3 proportionate review subcommittee of NRES.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.