The study was approved by the Hunter New England Human Research Ethics Committee (reference number: 2020/ETH01726).
- Participant selection
The John Hunter Children’s Hospital and the Children’s Hospital at Westmead medical records database between 2010-2020 were accessed to create an eligibility list for groups 1 and 2 below.
Group 1 (consented for repair) consisted of parents of children operated on for hypospadias (distal or proximal).
Group 2 (declined repair) included parents who consulted a surgeon for their son’s hypospadias and there is no record of a subsequent repair.
Group 3 (control) included parents whose children had a non-medically indicated circumcision. This list was developed from a surgeon’s (GS) private practice database (also between 2010 and 2020) since this procedure is no longer provided in public hospitals in Australia.
2. Sample size calculation
The incidence of surgical complications is correlated with DR [4, 5]. The reported complication rate associated with neonatal and infantile circumcision is around 1.5% [6]. It was not possible to gather data for this group on other factors believed to be associated with postoperative parental DR (parental decisional conflict, education level etc). We made a conservative estimate that the prevalence of parental DR in the elective ‘non-medical’ circumcision group was unlikely to exceed 10%.
We expected this control group to be much smaller than the hypospadias group (10 times smaller). We allowed for a low response rate as previous similar studies have a response rate of 37.4% [4]. We also allowed for a further 10% attrition due to incomplete responses. Using power parameters of α=0.05 and β=0.2, these assumptions produced a required sample size of 234 invitees for the repair group, and 34 for the control group. All parents who declined the repair were invited to participate in the survey.
3. Survey Design
The online anonymous survey was designed and administered using the secure web application tool REDCap, hosted at the Hunter Medical Research Institute (Newcastle, Australia) [7, 8]. After the introductory page providing the study information, and an electronic consent form, the questionnaire was specific to the group to which the respondent belonged. In addition to questions specific to their group, there were questions common to all three cohorts (See Additional file 1 for a full list of survey questions).
The construct of the survey is explained below.
3.1. Decision Regret assessment tool
The previously validated tool, the Decision Regret Scale (DRS), was used to assess Decision Regret. This tool has been consistently used in studies of parental DR in hypospadias [3, 9]. There are no rigid cut points for the severity of DR using this scale [10], however, previous studies have classified DR as no DR (0), mild (1-25) and moderate-to-severe (>25). We have followed this convention, but have also reported the numerical values (median, IQR, mean, SD).
3.2. Parental Perspectives and Open-Ended Questions
A series of three questions, specific to a hypospadias repair, was asked about the reasons parents opted for surgery, and the most and the least satisfying outcomes of the repair. These questions had multiple suggested answers as well as a free text option.
The parents from both hypospadias groups were offered a question about their preference regarding circumcision to explore if their views on completely elective surgery might influence DR. Parents were also asked whether they felt adequately supported in their decision, and whether there could have been anything to have helped them to be more content with the decision they made. Those questions also had a free-text option.
3.3 The Index of Relative Socio-economic Advantage and Disadvantage
The parents were offered to indicate the postcode they resided when the decision for their son was made. The postcode area then was located on the interactive map of the Australian Bureau of Statistic’s Index of Relative Socio-economic Advantage and Disadvantage (IRSAD) [11]. We used the latest index published in 2011 which is based on the information from the five-yearly Census.
3.4. Demographics
Respondents were asked about their age, gender, and highest completed level of education.
4. Survey distribution
Parents were invited using an email (where electronic medical records provided an email address) or a text message from a phone number specific to the study. Parents who had neither email nor mobile phone number in the record were excluded from the study.
A total of three rounds of email invitations were sent two weeks apart. Parents with an invalid email address and/or only a mobile number available were texted once only. After that the access to the questionnaire was closed and the hyperlinks inactivated.
5. Statistical analyses
Descriptive statistics were utilised to characterise the groups, and the prevalence and severity of DR were calculated. A non-parametric test of significance (Mann-Whitney U test) was used to explore the difference in DR score between the groups, as well as the difference in IRSAD. The Chi-square test was applied to assess the difference in prevalence of DR and the effect of circumcision preference on parental decision.
For the qualitative analyses, the free-text answers were grouped into four categories based on whether parents had DR or not and whether their child had the repair or not, and they were uploaded into NVivo (released in March 2020) [12]. Using an inductive method, all mentioned references were coded and the number of mentions was counted [13].