Journal of Rehabilitation Medicine 51-5 | Page 38

354 M. McGrath et al. we completed a systematic review and meta-synthesis of qua- litative research regarding experiences of stroke survivors and their partners in relation to post-stroke sexuality (1). Secondly, a content review of existing sexual rehabilitation programmes for adults following acquired disability was completed. Thirdly, we examined the curricula of published interventions, which addressed sexuality after stroke (16–18). A list of potential content was then organized into the 6 broad categories listed above and presented to participants in the first Delphi survey. For the first Delphi survey, participants were initially asked to rate the value of each content area in relation to sexual reha- bilitation programmes, using a 5-point-Likert scale. Secondly, participants were asked to identify and rank the 10 content areas they deemed to be most important for inclusion in a sexual rehabilitation programme, from 1 (highest) to 10 (lowest) in importance. Next, participants were asked to propose additional content not covered in the survey, responding to an open-ended question. The same process was completed for timing of a sexual rehabilitation programme, method of programme delivery and involvement of specific disciplines. For participants with self-reported communication difficul- ties, the aphasia-friendly version was modified using aphasia- friendly principles (24) with pictographic support, simplified language, bolding and larger font. Pictographic supports for sexuality were used with permission from the Aphasia Institute Canada (27). Three participants required the modified version, with the remainder (n = 5) completing the online survey with support from the speech pathologist as required. The online survey was formatted so that participants could return and complete it later; some people with mild self-reported com- munication difficulties reported that this was helpful, as they could complete the survey more slowly with breaks. Round 2. In round 2, participants were presented with content areas from round 1, along with a summary of responses obtained from round 1. Participants were asked to rate the importance of each content area, preferred methods of delivery, timing of delivery and discipline using the same 5-point Likert scale. Finally, participants were asked to reflect on results from round 1 and provide feedback on aggregated responses via a series of open-ended questions. Analysis Analysis was conducted for the overall sample. Descriptive statistics were used to calculate responses about content area, timing, methods and disciplines that the overall group deemed most important. For each content area, the mean importance score, the 25 th and 75 th percentile and interquartile range were calculated. To determine the degree of consensus around the importance score, a score was calculated for each content area by dividing the interquartile range by 2. Following guidelines proposed by Burnette and colleagues (28), a consensus score of < 1 indicated high consensus, a score of 1 indicated moderate consensus, and a score > 1 indicated low consensus. Two approaches were used to identify the content areas, timing, methods and disciplines deemed of highest importance by respondents. Firstly, we considered the mean priority score given to each content area. Scores ≤ 2.4 were categorized as low priority, scores ≥ 2.41 and scores ≤ 2.8 as intermediate priority, and scores ≥ 2.81 categorized as high priority. Secondly, the percentage of respondents who included a content area in their top 10 priorities was calculated. After reviewing the spread of scores, categories were assigned as follows: > 25% = high prio- rity, 15–25% = intermediate priority, and < 15% = low priority.   www.medicaljournals.se/jrm Finally, the 2 priority ratings (mean priority score and per- centage of respondents including the content area in their top 10 priorities) were synthesized into 1 priority code with 3 tiers. Tier 1 content, timing, delivery methods and disciplines were rated as either: (i) high by both rating schemes; or (ii) high by one rating scheme and intermediate by the second rating scheme. Tier 2 content were rated as either: (i) intermediate by both rating schemes or (ii) high by one rating scheme and low by the second rating scheme. Tier 3 content areas were rated as either: (i) low by both rating schemes or (ii) low by one rating scheme and intermediate by the second rating scheme. For example, the statement “Resuming sexual activity after stroke”’ had a mean priority score of 3.51 (high priority), with 57.4% of respondents identifying the statement as a top 10 priority area (> 25%, high priority), thus was included as a tier 1 content area. On the other hand, the statement “What is sexuality?” had a mean priority score of 3.57 (high priority) with 9.6% of respondents identify- ing the content area as a top 10 priority area (low priority), and thus was included as a tier 2 statement. A summary matrix was compiled to display the content, timing, delivery methods and disciplines according to consensus and priority. See supplemental Tables SI 1 , SII 1 , SIII 1 and SIV 1 for full information. RESULTS Participants The final sample size and characteristics are presented in Table I. Of the 102 participants who completed the first survey, 49% (n = 50) completed the second survey. For participants with self-reported communication difficulties (n = 8), 1 person (male) had a mild activity limitation on the AusTOMS (26) cognitive-commu- nication scale, 3 people (2 females and 1 male) had mild receptive and expressive language limitation on the AusTOMS language scale (26), and all elected to complete the online survey. One participant (female) had a moderate expressive language limitation, a mild receptive language limitation, and competed the online survey. Another participant (male) had moderate ex- pressive and receptive language limitations and chose to complete the aphasia-friendly survey. Finally, 2 male participants had moderate/severe language and speech limitations (apraxia) and completed the aphasia- friendly version with supported communication. Due to time constraints associated with data collection the 3 participants who completed the aphasia-friendly version of the survey were unable to participate in the second round of data collection. Content for inclusion in sexual rehabilitation following stroke: priorities and consensus During round 1, participants were presented with 47 potential content areas for inclusion in sexual rehabilita- http://www.medicaljournals.se/jrm/content/?doi=10.2340/16501977-2548 1