Skip to main content

Advertisement

Table 8 Findings from quality appraisal

From: Getting under the skin of the primary care consultation using video stimulated recall: a systematic review

First author and year Sampling and consent Effect of video or study methods on behaviour Other methodological issues identified from QA using CASP tool
Ali [20] No mention. States GPs were recorded over a period of time to try and reduce effect Mentions inclusion criteria but doesn’t describe these. Not clear in interview if interpreter was used or not, and what questions the patient was asked. Analysis not clearly described. Conclusions appear to be derived from literature review rather than empirical findings.
Characteristics of consenters described in unreferenced related paper only
Als [21] States attempted to recruit a sample of variation, characteristics and consent not described No mention Analysis not described in detail.
Arborelius [17, 2228], Characteristics of consenting patients described but not non-consenters. Mentions in 2 papers the influence of the camera was minimal (self-report from participants) Participant comments during VSR often not aligned to research question as only neutral prompts, therefore small number of comments relevant to study aims [23, 24].
Research question not aligned to sampling resulting in small numbers of relevant consultations for some papers [17, 25]. Analysis clearly described in 2 papers in this group [23, 27].
Possible over-interpretation of participants’ comments (particularly assumptions on when GP had failed to ‘grasp’ situation) [25, 27] with limited discussion of implication of findings [24]
Analysis mostly conducted across case and not within case: within cases analysis and comparison may have enhanced analysis and understanding of cases where difficulties exist in the consultation [26] (where within case approach was used, only 1 minute of consultation analysed [23]).
Blakeman [16, 29] Characteristics of consenting patients and GPs described but not non-consenters. No mention Data collection, rationale for study and analysis described in detail. Possible limited conclusions to be drawn from the study of one consultation when studying self-management support which may happen longitudinally in the doctor patient relationship.
Only empirical quotes from nurses reported in 2nd paper, yet conclusions refer to doctors and nurses. In 2nd paper, no discussion about how context of nurse or doctor consultation would influence findings in relation to QOF.
Bugge [19] Characteristics of consenting patients described but not non-consenters. Limited characteristics of GPs described Brief mention as limitation Relative contribution of different post consultation interviews not described (3 per participant).
Analysis well described.
Cegala [30] Characteristics of consenting patients and GPs described but not non-consenters. No mention. Effect on behaviour may be more likely as consultation taken out of normal surgery context and separate microphone on table. Paper based on assumption that participant’s spontaneous comments during playback (with no guided prompts) can be used to draw conclusions about patient perceptions of doctor competence in communication exchange.
No information about sampling. No empirical quotes to support findings.
Coleman [14] Characteristics of consenters and non-consenters presented. GPs sampled to represent a range of attitudes to smoking Discussed as potential limitation. Quantitative methods to support sampling helped gain a maximum variation sample.
Analysis well described.
Author’s role as GP and peer to GP participant’s not explored.
Cromarty [31] No mention of details of video selection or recruitment (videos selected by participating GPs and not researcher) No mention Relative contribution of different phases of post consultation interview not described (unprompted, with video recall and then written transcript).
Analysis not described in depth.
Epstein [15] Characteristics of consenting patients and GPs described but not non-consenters. One comment that GPs stated not affected. Robust analysis strengthened by different approaches including coding of behaviours, attention to conversation flow and classification scheme of the level and depth of discussion of HIV risk.
Discussion of how GPs volunteering to be video recorded may not be representative of GP population. More than one consultation per GP facilitated robust analysis.
Purposive sampling used to identify patients/ consultations more likely to contain discussion of HIV risk Not clear how video shown or VSR procedure.
Frankel [32] No mention No mention Research question or theoretical framework lacking.
Sample size unclear Participant comments (GP or patient) on video not confidential and revealed to other participant. Consent not mentioned.
Gao [33] Characteristics of consenting patients described but not non-consenters. Limited characteristics of GPs described No mention Recruitment strategy not entirely appropriate: GP interviews not needed to answer research question and weren’t utilised.
Three stage analysis clearly described.
Henry [18] Variation sampling of patients to gain mix of gender, age and race. GPs sampled with respect to years in practice and specialty No mention Insufficient detail about structure of interview or VSR procedure to judge how appropriate study method was for exploring tacit clues.
No discussion of how context of health maintenance consultations might influence findings.
Rosenburg [34, 35] Characteristics of sample described (patients and interpreters), but not non-consenters No mention Conclusion not supported by results and patient views would have added value and been relevant to research question [34].
Little information about VSR procedure of format of interview [35].
Rosenburg [13] Recruitment well described. Characteristics of sample described, but unclear how many underwent VSR No mention Method successful in identifying consultations of interest and evidence supports authors’ conclusions. No discussions of limitations.
Patients made few comments over video and structure of interview not clear.
Saba [36] Characteristics of sample described but low consent rate not discussed. Brief mention of possible effect Robust analysis strengthened by different approaches including analysis within and across cases, contrasting observed and subjective experiences of shared decision making to construct typology of SFM archetypes and using themes from interviews.
Timpka [37] Characteristics of consenting patients described but not non-consenters. Brief mention of possible effect Complex study but not clear how much video the participants viewed, the instructions the participants were given when watching the video or the consent arrangements.
Conclusion not supported by results.
Treichler [38] Case study of one patient. No mention of sampling. No mention Limitations associated with the study of one consultation.
Ventres [39, 40] Not described Brief mention Analysis well described but no empirical quotes to support findings. More description of consultation context would have increased credibility of findings.