The moral limits of medical research with children
Westra, A.E.
Citation
Westra, A. E. (2011, June 30). The moral limits of medical research with children. Retrieved from https://hdl.handle.net/1887/17752
Version: Corrected Publisher’s Version
License: Licence agreement concerning inclusion of doctoral thesis in the Institutional Repository of the University of Leiden
Downloaded from: https://hdl.handle.net/1887/17752
Note: To cite this publication please use the final published version (if applicable).
ChApter 5
perceived discomfort levels in healthy
children participatinG in vaccine research
CHAPTER 5 62
ABSTRACT
Objective. When assessing the risks of a research protocol, review boards need not only consider the possible harms, but also the expected discomfort levels caused by the various study procedures. However, data on how children experience various study procedures are scarce. This study aimed to assess perceived discomfort levels in healthy children aged 0-2 years undergoing vaccinations, venipunctures and nasopharyngeal swab taking.
Patients and Methods. Parents of infants participating in a trial investigating the effects of reduced dose pneumococcal conjugate vaccine schedules were asked to complete a self-administered questionnaire. The questionnaire assessed demographic characteristics, perceived discomfort levels in their children, perceived discomfort levels as compared to the parents’ expectations, and general views on participating in a trial.
Results. Of the 1005 children participating in the vaccine trial, 986 (98%) children completed the follow-up. Of these, 670 (68%) parents returned the self-administered questionnaire. The proportion of children with moderate or high levels of discomfort ranged from 12% for vaccination and 28% for nasopharyngeal swabbing to 49% for venipunctures. Within the reported age group of 0-2 years, increasing age was related with higher discomfort levels. In a majority of cases for all study procedures, the perceived levels of discomfort met the parents’ expectations.
Conclusion. This study shows that administering vaccinations and collecting nasopharyngeal and blood samples in young children in a research setting are generally well tolerated. The results may serve as a reference for study design and study protocol assessment of future pediatric trials.
a. e. westra1, e. van Gils2,3, f. aarts1, G.d. rodenburg2, r.h. veenhoven3, e. hak4, m. scharloo5, r. n. sukhai1, J. m. wit1, i. d. de beaufort6, e. a. m.
sanders2
1. Department of Pediatrics, Leiden University Medical Center
2. Department of Pediatric Immunology and Infectious Diseases, Wilhelmina Children’s Hospital, University Medical Center Utrecht
3. Department of Pediatrics, Spaarne Hospital Hoofddorp
4. Department of Pharmacy, Unit of PharmacoEpidemiology & PharmacoEconomics (PE2), University of Groningen
5. Department of Medical Psychology, Leiden University Medical Center 6. Department of Medical Ethics and Philosophy of Medicine, Erasmus Medical
Center Rotterdam, Submitted
INTRODUCTION
Over the last years, both the United States and the European Union have taken measures to stimulate pediatric drug and vaccine research.49-51 Such initiatives although important have made optimizing the system that protects children involved in such research all the more pressing. One issue requiring improvement is the risk-benefit assessment of proposed studies. In case of studies that offer the prospect of direct benefits to the subjects, review boards must evaluate whether these possible benefits outweigh the risks; in case of studies without possible direct benefits to the subjects, the risks must be reasonable in relation to the knowledge to be gained, and the risks may not be more than ‘minimal’
(or sometimes ‘a minor increase over minimal’).18 For both type of protocols, review boards need to assess how substantial the research risks actually are for the children at stake.
When assessing research risks, review boards need not only consider the possible harms, but also the expected discomfort levels caused by the various study procedures. This is reflected in the US Federal Regulations’ definition of
‘minimal risk’: ‘the probability and magnitude of harm or discomfort anticipated in the research are not greater in and of themselves than those ordinarily encountered in daily life or during the performance of routine physical or psychological examinations or tests.’ The World Medical Organization and the Council of Europe even explicitly distinguish the expected discomforts from the risks of serious harm and speak of ‘risks and burdens’ (the term ‘burden’ referring to the expected discomforts).25;32
When evaluating expected discomfort levels, review boards currently may rely on the experiences of experts, as review boards nowadays are required to include or consult an expert in the field of pediatrics.19;82 This is not yet an optimal situation, because the pediatric experts’ view will often originate from a rather clinical perspective, ignoring the fact that in case of research there is no medical reason justifying the procedure, while the experts’ view also may vary from person- to-person. Ideally, review boards should base their judgment on group-level data about how children experience various study procedures. Unfortunately though, such group-level data are rather scarce, especially when taking into account the potential differences between children in various age groups and in various health conditions.
This study aimed to assess perceived discomfort levels in healthy children aged 0-2 years participating in a pneumococcal vaccine trial and undergoing vaccinations, venipunctures and nasopharyngeal swabbing. The data on these three routine vaccine study procedures may serve as a reference for the design and ethical review of future pediatric trials.
CHAPTER 5 64
PATIENTS AND METHODS
Study setting and population
A randomized controlled trial investigating the effects of reduced dose pneumococcal conjugate vaccine (PCV-7) schedules on nasopharyngeal bacterial carriage in infants and household contacts during the first two years of life was performed in the Western part of the Netherlands from July 2005 until February 2008 (NCT00189020).96 After written informed consent had been obtained from both parents or guardians, infants were randomly allocated during the first home- visit into one of the three study groups to receive (a) PCV-7 at the age of 2 and 4 months (2-dose group), (b) PCV-7 at the age of 2, 4 and 11 months (2+1-dose group) or (c) none (control group). Children in the control group were offered a PCV-7 vaccination free of charge at the age of 24 months after completing the study. Parents were aware of the child’s vaccine schedule.
Vaccinations were intramuscularly administered in the leg or upper arm during regular well baby-clinic visits, together with routine immunization(s) according to the Dutch National Immunization Program. All other study procedures were performed during home-visits. During the first home-visit at the age of 6 weeks and during the four follow-up visits at 6, 12, 18 and 24 months of age, a nasopharyngeal sample was obtained from the children and at 12 and 24 months also from one of their siblings (if any). Deep nasopharyngeal swabs were taken transnasally with a flexible, sterile, dry cotton-wool swab by trained study nurses according to WHO standard procedures.97 Blood samples were obtained via venipuncture from a voluntary subgroup of children per study group at the age of 12 and 24 months (n»80 in both vaccine groups; n»30 in control group). A local anesthetic (Eutectic Mixture Local Anesthetic) was applied approximately 30-60 minutes before the venipuncture. All study procedures during home-visits were performed by trained study personnel.
Parents received no financial compensation for participation. An acknowledged Dutch review board (Stichting Therapeutische Evaluatie Geneesmiddelen (STEG), www.stegmetc.org) approved the study protocol.
Questionnaire
Approximately six to nine months after completion of the study procedures the parents of all subjects were asked to complete an anonymous self-administered questionnaire that was developed specifically for this study. The questionnaire assessed (1) demographic characteristics; (2) perceived discomfort levels in their children; (3) perceived discomfort levels as compared to the parents’ expectations;
(4) general views on participating in a vaccine trial. Questions regarding the perceived discomfort levels were based on a 4-point scale (‘none’; ‘mild’; ‘moderate’; ‘high’).
Questions regarding parents’ expectations were assessed using a three point scale (‘lower than expected’; ‘as expected’; ‘higher than expected’). Parents could
indicate with ‘yes’, ‘maybe’ or ‘no’ whether several design modifications would have changed their minds towards refusing to participate in the trial. Separate ethical approval for the questionnaire was received from the STEG.
Statistical analysis
Completed questionnaires were coded and analyzed using SPSS 17.0 for Windows.
For convenience of comparison, perceived discomfort levels were categorized as either ‘none or mild’ or ‘moderate or high’. The analysis of proportions consisted of constructing binomial point estimates and exact binomial 95%
confidence intervals (95CI) for each group. For statistical reasons, data regarding venipunctures at 12 and 24 months were combined. Differences in discomfort levels between study groups were tested using Chi-square-tests. Analysis of continuous variables consisted of point estimates and 95CI for means. Binomial logistic regression was used to examine the association between education level and perceived discomfort levels.
RESULTS
Response and subject characteristics
A total of 1005 infants were enrolled in the trial and 986 children (98%) completed the study follow-up. The questionnaire was administered to the parents of all 986 children who had completed the trial. Sixty-eight percent (n=671) of the parents completed and returned the questionnaire. The response rate varied among the three study groups: 60% in the 2-dose group, 78% in the 2+1-dose group and 66% in the control group. The majority of respondents (82%) had received tertiary education, of which 56% had finished vocational higher education or university. Table 1 shows some general characteristics of the parents that filled in the questionnaires.
Perceived discomfort levels and parental expectations
Vaccinations were considered to cause a moderate or high level of discomfort in only 6.0% (95CI 4.2%-8.6%) of the children at the age of 2 months. The level increased significantly with age up to 9.8% (95CI 6.8%-14.1%) for the vaccination at the age of 11 months and 18.1% (95CI 15.2%-21.4%) at the age of 24 months.
(Figure 1) The nasopharyngeal swabs were considered to cause a moderate or high level of discomfort in 20.4% (95CI 17.5%-23.7%) of the children during the first half year of life, which increased significantly with age to 28.7% (95CI 25.8%- 32.7%) around the first birthday and to 35.8% (95CI 32.2%-39.5%) around the second birthday. In (older) siblings, the proportion of children experiencing a moderate or high level of discomfort due to the nasopharyngeal swab was 44.5%
(95CI 39.1%-50.0%). Venipuncture was considered to cause a moderate or high level of discomfort in 49.3% (95CI 45.3%-53.3%).
CHAPTER 5 66
Table 1. Characteristics of parents who returned study questionnaire (n=671).
n %
Assigned vaccine schedule
2-dose PCV7-group a 199 30
2+1-dose PCV7-group b 260 39
Control group c 212 32
Questionnaire filled in by
Father 47 7
Mother 618 92
Other guardian or unknown 6 1
Level of highest finished education
Primary 4 1
Secondary 106 16
Tertiary
Community college 177 26
Vocational higher education 213 32
University 161 24
Unknown 10 1
First born child 345 51
PCV-7: 7-valent pneumococcal conjugate vaccine
a Receiving PCV-7 at age 2 and 4 months
b Receiving PCV-7 at age 2, 4 and 11 months
c Receiving no PCV-7
The study group to which the child was randomly allocated (being either one of the vaccine groups or the control group) did not significantly affect the perceived discomfort levels except for a higher proportion of children experiencing a moderate or high level of discomfort after the nasopharyngeal swab at the age of 12 months in children having received a booster vaccination 1 month earlier (32.6%, 95CI 27.1%-38.5% in the 2+1-dose group) compared to the unvaccinated control group (24.0%, 95CI 18.7% – 30.3%; p= 0.04). The parents’ level of education did also not significantly affect the perceived discomfort levels, nor did the collection of a blood and nasopharyngeal sample during the same visit.
97.8% (95CI 96.3%-98.6%) of the parents indicated that the perceived level of discomfort caused by vaccination was as expected or even lower than expected.
Corresponding figures for nasopharyngeal swabbing and venipuncture were 78.5% (95CI 75.2%-81.4%) and 65.9% (95CI 60.9%-70.5%) respectively. (Table 2)
General views on participating in the trial
Overall, parents were satisfied with their (children’s) participation in the trial. Their average rating of the trial was 8.3 (standard deviation (SD): 0.7) out of 10. Only 1% (95CI 0%-2%) of the parents indicated that in retrospect, they would rather not have made the decision to participate. Importantly, 82% (95CI 79%-85%)
79
Figure 1. Perceived discomfort levels in healthy children participating in vaccine research.
0 10 20 30 40 50 60 70 80 90 100
2 or 4 mo 11 mo 24 mo 2 or 6 mo 12 mo 24 mo sibling 12 or 24 mo
Vaccination Nasopharyngeal swab Venepuncture
Percentage
Study procedure
High Moderate Mild None
Figure 1. Perceived discomfort levels in healthy children participating in vaccine research.
Table 2. Perceived levels of discomfort as compared to parents’ expectations.
Na As expected Not as expected
n(%) Less discomfort
n (%) More discomfort
n (%)
Vaccination 669 468 (70) 186 (28) 15 (2)
Venipuncture 375 177 (47) 70 (19) 128 (34)
Nasopharyngeal swab
Child 664 399 (60) 122 (18) 143 (22) Sibling 315 196 (62) 30 (10) 89 (28)
answered that they would surely participate if asked again. Out of the possible study design modifications offered, only hospital-visits instead of home-visits would have surely (48%, 95CI 44%-52%) or maybe (another 40%, 95CI 37%-44%) changed their minds towards refusing to participate in the trial. (Table 3)
DISCUSSION
This study shows that the majority of the study procedures were perceived as causing no or mild discomfort, which met the parents’ expectations. Among the study procedures, venipuncture was most often considered to cause a moderate or high level of discomfort (49%), followed by the nasopharyngeal swabs and vaccination (overall, respectively: 28% and 12%). Within the reported age group,
CHAPTER 5 68
increasing age was related with higher perceived discomfort levels. This finding is in line with previous procedural distress studies reporting that age accounted for an important amount of variability in distress scores.93;98;99 In these studies younger instead of older children experienced more distress, but children in the youngest age groups were not younger than 2.5 years, suggesting that in particular toddlers are at risk of negative experiences. This may well reflect the developing awareness of children at this age. Furthermore, in this study, the children’s previous experiences with the procedures may have played a role due to the repetitive exposure to some of the study procedures. In addition, the older children within the age group of this study are more capable of showing some resistance, which may also have influenced the parents’ perceptions.
Regarding the venipunctures, our finding that approximately half of these were experienced as causing a moderate or high level of discomfort seems in line with a study by Davies et al., who reported 60% of 866 children aged 1.5 to 4.5 years being upset during/after venipuncture, and with a paper by Fradet et al., who reported that 36-64% aged 3 to 6 years old showed moderate to severe distress.99;100 Humphrey and colleagues found higher levels of distress up to 83% in a small sample of 70 children with varying age from 2.5 to 6 years.93 However, in contrast to questionnaires on parents’ perceptions as used in our and the previous mentioned studies, this study used the Groningen Distress Scale (GDS). The GDS is a behavioral observation scale based on crying, muscle tension and breathing, which has to be assessed by a trained observer. Serious distress is coded (according to the modified version) in case of a rigid posture, crying for less than 30 seconds (more than 30 seconds means ‘severe’ distress) and vocalization of pain.98 It may be that parents being familiar with their child’s behavior classify the same reaction as ‘mild’. This may also explain our finding that only 6-18% of vaccinations were experienced as causing moderate or high levels of discomfort in contrast with a previous study by Jacobson et al., who also used the GDS and observed serious distress or more in 81% of 150 children aged 15-18 months
Table 3. What would have changed the parents’ minds to reject the trial?
N No
n (%) Maybe
n (%) Yes n (%) Vaccination already available through national
immunization program a 668 286 (43) 313 (47) 69 (10)
Different research nurses at each house-visit 667 475 (71) 132 (20) 60 (9) Working >3 days a week b 456 292 (64) 135 (30) 29 (6) More questionnaires to fill in 666 419 (63) 202 (30) 45 (7) Hospital-visits instead of home-visits 664 78 (12) 268 (40) 318 (48)
a PCV-7 was introduced in the Dutch nation infant immunization program for all infants born after March 31, 2006; trial enrolment was finished before this date.
b only available for those parents who did not work >3 days a week
undergoing vaccination.98 Another explanation for the lower scores may be that in our study all procedures were performed during home-visits, that is, in a familiar environment, by people without white coats. We have not investigated whether the home-visits have positively influenced the (perceived) discomfort levels of the three procedures at stake, but from the average rating of the trial (8.3 out of 10) as well as from the low drop-out number, one can at least conclude that this approach was much appreciated.
Two limitations of this study should be recognized. First, the study was conducted in a very young population making it impossible to interview the subjects themselves. Parents were regarded as the most suitable surrogates since they know their child’s behavior best. Second, the questionnaire was voluntary, resulting in a 68% response rate. This may have introduced some response bias.
The response rate in the control group did not differ from that in the vaccine groups, but considering the high percentage of highly educated parents in our sample compared to the Dutch population (56% versus 27%; www.cbs.nl) it is for instance not unlikely that the group of non-responders was less well educated.
Unfortunately, due to the anonymous character of the questionnaire, information on the non-responders is not available.
This exploratory study may serve as a reference for the design and ethical review of future pediatric trials. The data need careful interpretation though, in particular in relation to the concept of ‘minimal’. Clearly, in defining groups at risk for high levels of discomfort, age should be an important factor. However, none of the procedures was experienced in the same way by all children, even not within a particular age group. This means that the question of whether the expected level of discomfort of a procedure (in a certain age group) can be considered ‘minimal’
or not is not easily answered. Wherever one draws the line, there may always be children included that will experience moderate or high levels of discomfort.
We suggest that review boards assess, on the basis of group-level data such as presented in this study, whether the expected level of discomfort of a proposed study can be considered acceptable for a certain majority of potential subjects.
For the fine-tuning, when deciding whether or not to enroll an individual child in a study, active participation of parents and researchers is indispensable: they are in the best position to assess whether the adjective ‘minimal’ will also apply to the individual child in front of them.
CONCLUSION
This study shows that administering vaccinations and collecting nasopharyngeal and blood samples in young children in a research setting are generally well tolerated. In order to improve and objectify the ethical review process, more studies collecting group-level data are necessary.
