Developing outcome measures assessing wound management and patient experience: a mixed methods study

Objectives To develop outcome measures to assess practical management of primary surgical wounds and patient experience. Design Mixed methods, including qualitative interviews and data extraction from published randomised controlled trials (RCTs). Setting Two university-teaching NHS hospitals and three district NHS hospitals in the South West and Midlands regions of England. Participants Sixty-four patients and 15 healthcare professionals from abdominal general surgical specialities and obstetrics (caesarean section). Methods Measures were developed according to standard guidelines to identify issues relevant to patients’ experiences of surgical wounds and dressings, including analysis of existing RCT outcomes and semi-structured interviews. These were written into provisional questionnaire items for a single outcome measure. Cognitive interviews with patients and healthcare professionals assessed face validity, acceptability and relevance. Findings from interviews were regularly shared with the study team who suggested amendments to modify and reword items to improve understanding before further iterative testing with patients and healthcare professionals. Results Analyses of existing RCT outcomes and interviews produced a total of 69 issues. Pretesting and iterative revision established the need for two separate measures. One measure addresses healthcare professionals’ experience of wound management in two key areas: exudate and its impact, and allergic reactions to the dressing. The other measure addresses patients’ experience of wounds in seven key areas: wound comfort, dressing removal, dressings to protect the wound, impact on daily activities, ease of movement, anxiety about the wound and satisfaction with dressing. Each measure took less than five min to complete and both were understood and acceptable to patients and healthcare professionals. Conclusion This in-depth study has developed two measures to assess practical management of primary surgical wounds and patient experience. Further work to test their validity, reliability and application to other settings is now required. Trial registration number HTA - 12/200/04; Pre-results.

• This study used robust methods to identify key issues of outcome that should be assessed in future randomised controlled trials of wound dressings. Interviews provided a rich account of the key factors that affected wound management and patient experience while a purposeful sampling strategy ensured that perspectives were captured from a range of participants. Data produced from the interviews were supplemented by an extensive literature search to ensure a comprehensive list of issues was considered.
• Future work is needed to test the reliability, validity and sensitivity to change of the new measures.  [1]. It is common practice to apply dressings over the closed wound in adult surgery and many different dressing types are available [2]. Whilst the role of dressings in covering primary surgical wounds in reduction of rates surgical site infection (SSI) is unclear, it is recognised that they are important for patient comfort and management of wound exudate in the early post-operative period [3].
A recent Cochrane systematic review summarised data relating to wound dressings and risk of SSI in primary surgical wounds. No evidence was found to suggest that any type of dressing significantly reduced the risk of developing an SSI compared with leaving wounds exposed; neither was there any benefit associated with particular dressings [2]. The authors concluded that decision-making around dressings may need to be informed by costs and practical issues. The review highlighted that measures for assessment of wound cosmesis (in the longer term) are available [4,5]. It reported, however, that measurement of wound issues such as exudate and patient experience is difficult because of a lack of well-developed and validated measurement tools.
In the past, outcome measures were often derived from clinicians' views, rather than patients', on issues of importance [6]. However, more recently, a patient-centred approach has been advocated [7]. This has resulted in the increasing use of patient-reported outcome measures (PROMs) in clinical trials and usual clinical practice [8]. The development of PROMs may include use of qualitative research methods that provide the opportunity to elicit and characterise patients' experiences of their health conditions and treatment.
Qualitative methods can also define health care professionals' experiences of care and management. [9], which can be supplemented by expert input and studies in published literature [10]. This paper describes the development of measures to assess practical wound management issues, symptoms and patient experience associated with primary surgical wounds.

Study design
Measures were developed according an existing framework for developing PROMs [11,12], also incorporating guidance on eliciting health domain concepts using qualitative methodologies [9,10]. The study is reported according to qualitative reporting guidelines Phase 1: Generation of relevant issues

Interviews
Participants were recruited as part of a wider feasibility study to explore whether a trial comparing different types of dressings, and dressing versus no dressings, is possible (The Bluebelle study: a feasibility study of three wound dressing strategies in eLective and unplanned surgery, HTA -12/200/04 [14,15]. Eligible patients had undergone abdominal general surgery or caesarean section were identified and approached by research nurses and surgical trainees (NB, DM). DE, CMM and LR contacted interested patients to arrange interviews. A purposeful sampling strategy ensured that perspectives were captured from a range of participants [10]. Within this sampling approach, maximum variation was sought in relation to age, gender, ethnicity, type of surgery, dressing type and location.
Interviews were conducted with patients to explore and characterise experiences of wounds and dressings by LR, DE and CMM. A topic guide was developed (based on literature and views of health care professionals in the Bluebelle study team) to ensure that discussions covered the same core issues but with sufficient flexibility to allow new issues of importance to the informants to emerge (See Additional File 3).
Interviews were audio-recorded and transcribed in full. Transcripts were imported into NVivo (version 10). All data relating to outcomes and issues of importance to patients that were relevant to dressing use and the practical management of the wound in the initial time period post-surgery were assigned labels (coded) by DE or CMM. Data were analysed using techniques of constant comparison derived from grounded theory methodology, and emerging codes across the dataset were then compared to look for shared or disparate views among participants [16]. A subset of approximately half of the interviews (n = 19) was double coded by a third researcher (LR) to highlight any differences in the interpretation of codes [9]. Data collection and analysis continued until the team were confident that saturation had been reached i.e. no more patterns or themes emerged from the data.

Literature search to identify existing tools
Three systematic reviews [17][18][19] were used to identify RCTs that included outcomes relating to wounds and dressings. Papers were scrutinised for outcomes relating to practical wound management or symptoms and patient experience. RM and JB extracted data including verbatim descriptions of these outcomes, and when and how they were measured.

Synthesis of literature and qualitative data
A list of issues from the analysis of the interviews and literature search was collated into an item tracking matrix [20]. DE, JB, LR, RM and CMM agreed on a set of words or phrases to reflect each issue and also noted additional phrasing made by participants in a subsequent column [9]. Issues which were conceptually similar were organised into categories. For instance, issues such as 'itchiness/irritation', 'presence of pulling sensation', and 'tightness of wound' were mapped into a 'wound comfort' category.
Phase 2: 'Operationalisation': Construction of a provisional measure DE, JB, LR, RM and CMM used the item tracking matrix to agree which issues should be written into questionnaire items. Items featured words and phrases used by patients in the interviews to enhance content validity [10,21]. Interviews explored the acceptability of the measure and coverage of patients and health care professionals' concerns (in terms of language, accuracy, and relevance) as well as layout [10]. During each interview, participants were asked to complete the measure by reading each item aloud and commenting on their understanding. Interviews were guided by a series of probes (e.g. 'What does this item mean to you?', 'Are there other ways you would describe it?'; [22]). Participants' body language was also observed and prompted further discussion about specific items (such as the participant nodding in agreement or frowning) [9]. A copy of the topic guide is available (See Additional File 3).
DE, CM and MMM maintained detailed field notes from each interview, describing suggestions for modifications and improvements to the provisional measures.
Operationalisation and modification of the measures was an iterative process. Findings from cognitive interviews and suggestions for amendments were regularly disseminated to the Bluebelle Trial Management group (TMG), which consisted of a multidisciplinary group of health care professionals, including surgeons, health services researchers and research nurses. Each stage of feedback informed amendments to modify and reword items to improve understanding, which was repeated following efforts to revise questions and eliminate problems [22]. This process continued until no new issues were identified and no further refinements were believed to be necessary.

Interviews
Semi-structured face-to-face or telephone interviews were conducted between July 2014 and

Literature search to identify existing tools
Twenty six studies that included outcomes relating to wounds and dressings were identified . Only two studies included a validated instrument to assess outcomes [25,39].
These were for long term scarring [4] and cosmesis [5]. However, no studies reported using validated measures relating to issues associated with practical wound management and patient experiences in the early post-operative period. Overall, papers provided very little detail on the rationale for including each outcome and how these outcomes had been measured. Attempts were made to contact the authors for more information, with only one reply.

Synthesis of literature and qualitative data
When describing experiences in the interviews, patients commented on several factors that affected perceptions of how well their wound was healing, including how it felt (tightness, pain, and itchiness) and whether any fluid had leaked from the wound. The literature review showed these issues had briefly been captured in some previous (unvalidated) outcome measures.
All patients had at least one dressing applied after surgery, although this varied between adhesive coverings (absorptive or non-absorptive) and tissue adhesive as-a-dressing ('tissue glue'). Both the interviews and existing outcomes from the literature highlighted that . The interviews also demonstrated that there were psychological factors that affected dressing experience and satisfaction (i.e. anxiety about cleanliness of the wound). In particular, patients in the interviews described potential benefits to having tissue glue as a dressing, including that it was transparent, waterproof, did not require multiple applications and came off naturally.
The interviews and the literature search produced a total of 69 issues. These were grouped into ten broad categories: wound comfort, exudate and its impact, allergic reactions to the dressing, dressing key removal, dressings to protect the wound, impact on daily activities, ease of movement, anxiety about the wound, satisfaction with dressing and wound appearance. Table 2 provides illustrative quotes for the categories identified.  Participants were recruited from two hospitals in two UK cities. Demographics are shown in Table 1.
Interviews highlighted issues with content in the initial measure. For example, items regarding the colour of the wound exudate were removed. Questions were rephrased to focus on the experience of having a dressing rather than general recovery after surgery (i.e. 'Have you been able to perform everyday tasks? (i.e. showering/bathing)' was changed to 'Has your dressing prevented you from showering/washing?'). Additionally, since four patients commented that the smell of their wound was missing on the measure, an item was added to capture this.
The measure had intended to be administered two days after surgery, although feedback suggested that this needed to be completed up to day four as the patient may be disorientated from surgery in the first few days. However, since there were clear differences in recovery with caesarean section and abdominal surgery patients, a timeframe of within four days of surgery was set, and the measures recorded the date of surgery and date completed to determine context of responses.
Feedback from patients suggested it was difficult to respond to questions about exudate, since a healthcare professional cared for their wound whilst they were in hospital. If their dressing had been changed, they were also uncertain about the reason why (i.e. simply as part of standard practice or for other reasons). Therefore, the study team decided to separate the Analyses of three systematic reviews [16][17][18] identified RCTs that included outcomes relating to wounds and dressings. Two validated tools were identified although neither are relevant in the early post-operative phase and both focussed on cosmesis and scar appearance [4,5]. No validated measures for assessing early wound issues or patient experience were identified. In addition, there was very little detail of how the items were developed and it was not apparent that patient input had been sought. True patient-centred outcome measures require full consideration of patients' experiences and views' [7,8].
The main strength of this research, therefore, is the use of qualitative research methods to provide important insights into the under researched area of early issues related primary surgical wounds relating to practical wound management and patient experience. Interviews provided a rich account of the key factors that affected wound recovery and satisfaction, and a purposeful sampling strategy ensured that perspectives were captured from a range of participants. To further enhance the reliability of the findings, the data were analysed by multiple researchers [9]. Data produced from the interviews were supplemented by an extensive literature search to ensure a comprehensive list of issues were initially generated, and therefore acted as a method of triangulation to increase the plausibility and dependability of the interview data [10].
It is important to note that these measures have only been pre-tested in relation to primary surgical wounds. Wounds that are intentionally left open or wounds that have developed problems are likely to require dressings that have advanced practical properties that are tailored to the wound requirements [14]. Although participants were purposefully sampled, most had had a dressing of some kind (94%). A prospective real-time survey of dressings has demonstrated that this reflects current practice [15]. In addition, these measures have only been pre-tested in relation to abdominal surgical wounds. However, characteristics of wound healing in this area are likely to be consistent with many other parts of the body. Furthermore, these measures focus specifically on the experience of dressings -methods of wound closure (i.e. potentially leading to differential ease of removal of sutures or staples) may also affect patient experience, although this would require further investigation.

Data and findings consistent
Was there consistency between the data presented and the findings?
There is consistency between the data presented and the measures developed. The item tracking matrix provides an overview of the key findings and how these were used to develop the initial measure (See additional file) 31 Clarity of major themes Were major themes clearly presented in the findings?
The themes are clearly presented in the were regularly shared with the study team who suggested amendments to modify and reword items to improve understanding before further iterative testing with patients and health care professionals.
Results: Analyses of existing RCT outcomes and interviews produced a total of 69 issues.
Pre-testing and iterative revision established the need for two separate measures. One measure addresses health care professionals' experience of wound management in two key areas: exudate and its impact, and allergic reactions to the dressing. The other measure addresses patients' experience of wounds in seven key areas: wound comfort, dressing removal, dressings to protect the wound, impact on daily activities, ease of movement, anxiety about the wound and satisfaction with dressing. Each measure took less than five minutes to complete and were understood and acceptable to patients and health care professionals.
Conclusion: This in-depth study has developed two measures to assess practical management of primary surgical wounds and patient experience. Further work to test their validity and reliability and application to other settings is now required.

STRENGTHS AND LIMITATIONS OF THIS STUDY
• This is the first study to explore the important issues related to the practical management of primary surgical wounds and patient experience immediately following surgery. • This study used robust methods to identify key issues of outcome that could be used to inform decision-making around dressings. Interviews provided a rich account of the key factors that affected wound management and patient experience while a purposeful sampling strategy ensured that perspectives were captured from a range of participants. Data produced from the interviews were supplemented by analysis of existing RCT outcomes to ensure a comprehensive list of issues was considered.
• Future work is needed to test the reliability, validity and sensitivity of the new measures.  (1). It is common practice to apply dressings over the closed wound in adult surgery and many different dressing types are available (2). A recent Cochrane systematic review summarised data relating to wound dressings and risk of surgical site infection (SSI) in primary surgical wounds. No evidence was found to suggest that any type of dressing significantly reduced the risk of developing an SSI compared with leaving wounds exposed; neither was there any benefit associated with particular dressings (2).
Decision-making around dressings may therefore need to be informed by other properties and qualities that dressings can offer, such as absorption of exudate, patient comfort, offering physical protection, facilitating wound observation, and meeting patients' desires for wound coverage (3). Whilst measures for assessment of wound cosmesis (in the longer term) are available (4,5), there is a lack of well-developed and validated measurement tools relating to practical wound management or patient experience (2,3,6). Such an instrument could be used to monitor the care of individual patients (e.g. assessing the ability of dressings to manage specific symptoms), audits (e.g. quality assurance) and research (e.g. comparing patient satisfaction).
The development of patient-reported outcome measures (PROMs) increasingly includes the use of qualitative research methods that provide the opportunity to elicit and characterise patients' experiences of their health conditions and treatment (7,8). Qualitative methods can also define health professionals' experiences of care and management (9). Data can be supplemented by expert input and studies in published literature (10,11). This article describes the development of measures to assess practical wound management issues, symptoms and patient experience associated with primary surgical wounds.

Study design
Measures were developed according an existing framework for developing PROMs (12,13), also incorporating guidance on eliciting health domain concepts using qualitative methodologies (10,11,14). The study is reported according to qualitative reporting guidelines (See Supplementary File 1). Phase 1 aimed to produce a comprehensive list of potential issues relating to wound and dressing experience and practical management issues.
Phase 2 developed issues identified from Phase 1 into questionnaire items. Phase 3 Interviews were audio-recorded and transcribed in full. Transcripts were imported into NVivo (version 10). All data relating to outcomes and issues of importance to patients that were relevant to dressing use and the practical management of the wound in the initial time period post-surgery were assigned labels (coded) by two experienced qualitative researchers.
Data were analysed using techniques of constant comparison derived from grounded theory methodology, and emerging codes across the dataset were then compared to look for shared or disparate views among participants (17). A subset of approximately half of the interviews (n = 19) was double coded by third experienced researcher to highlight any differences in the interpretation of codes (10). Data collection and analysis continued until the team were

Extraction of information from three systematic reviews
Three systematic reviews were used to identify RCTs which included outcomes relating to patient experience and management of wound healing. Published papers reporting the studies were obtained where possible. Relevant data were extracted on outcome (as described by the authors), verbatim wording to measure outcome, who reported the outcome, measurement scale and assessment time point. Attempts were made to contact the authors for more information.

Synthesis of findings from interviews and data extraction
A list of issues from the analysis of the interviews and analysis of existing RCT outcomes was collated into an item tracking matrix, in line with guidance for developing PROMs (19).
This is available in the online appendix (see Supplementary File 3). The study team agreed on a set of words or phrases to reflect each issue and also noted additional phrasing made by participants in a subsequent column (10). Issues which were conceptually similar were organised into categories. For instance, issues such as 'itchiness/irritation', 'presence of pulling sensation', and 'tightness of wound' were mapped into a 'wound comfort' category.
Phase 2: 'Operationalisation': Construction of a provisional measure The item tracking matrix was used to determine which issues should be written into questionnaire items. Items featured words and phrases used by patients in the interviews to enhance content validity (11,20). Cognitive interviews are used widely in questionnaire development (10) and involves asking respondents to verbalise their thoughts while answering questions (21). This methodology enabled us to explore the acceptability of the measure and coverage of patients and health care professionals' concerns (in terms of language, accuracy, and relevance) as well as layout (11). During each interview, participants were asked to complete the measure by reading each item aloud and commenting on their understanding. Interviews were guided by a series of probes (e.g. 'What does this item mean to you?', 'Are there other ways you would describe it?'; (21)). Participants' body language was also observed and prompted further discussion about specific items (such as the participant nodding in agreement or frowning) (10). A copy of the topic guide is available (See Supplementary File 2).
The qualitative team maintained detailed field notes from each interview, describing suggestions for modifications and improvements to the provisional measures.
Operationalisation and modification of the measures was an iterative process. Findings from the interviews and suggestions for amendments were regularly disseminated to the Bluebelle Study Group, which consisted of a multidisciplinary group of health care professionals, including surgeons, health services researchers and research nurses. Each stage of feedback informed amendments to modify and reword items to improve understanding, which was repeated following efforts to revise questions and eliminate problems (21). This process continued until no new issues were identified and no further refinements were believed to be necessary.

Interviews
A total of 39 interviews were conducted between July 2014 and July 2015. Interviews were conducted in person (n=10), unless patients preferred to be interviewed via telephone  Table 1.

Extraction of information from three systematic reviews
Published papers for twenty six studies that included outcomes relating to patient experience and management of wound healing were identified from the three systematic reviews . Only two studies included a validated instrument, or modification of a validated instrument, to assess outcomes (24,38). These were for long term scarring and cosmesis (4,5). However, no studies reported using validated measures relating to issues associated with practical wound management and patient experiences in the early post-operative period.
Descriptions of outcomes were heterogeneous and often poorly defined. The most common reported outcomes related broadly to cosmetic result (reported in 15/26 studies), dressing changes (e.g. frequency, comfort, ease of application and removal; reported in 11/26 studies), and skin reactions (e.g. itching, blistering; reported in 10/26 studies). Full data extraction from the 26 studies is included in Supplementary File 4.

Synthesis of findings from interviews and data extraction
When describing experiences in the interviews, patients commented on several factors that affected perceptions of how well their wound was healing, including how it felt (tightness, pain, and itchiness) and whether any fluid had leaked from the wound. Analysis of existing RCT outcomes showed these issues had been captured in some previous (unvalidated) outcomes. All patients had at least one dressing applied after surgery, although this varied between adhesive coverings (absorptive or non-absorptive) and tissue adhesive as-a-dressing ('tissue glue'). Both the interviews and the analysis of existing RCT outcomes highlighted that multiple practical advantages of dressing use (including ability to contain exudate and ease of removal). The interviews also demonstrated that there were psychological factors that affected dressing experience and satisfaction (i.e. anxiety about cleanliness of the wound).
Patients with tissue glue as a dressing commented that they had been surprised that their wounds had been dressed this way (rather than adhesive dressings which they had had in the past for other wounds). However, these patients stated that compared to past experiences of adhesive dressings, they liked how glue was transparent, waterproof, did not require multiple applications and came off naturally.
The interviews and the analysis of existing RCT outcomes produced a total of 69 issues.
These were grouped into ten broad categories: wound comfort, exudate and its impact, allergic reactions to the dressing, dressing removal, dressings to protect the wound, impact on daily activities, ease of movement, anxiety about the wound, satisfaction with dressing and wound appearance. Table 2 provides illustrative quotes for the categories identified. True patient-centred outcome measures require full consideration of patients' experiences and views' (7,8). The main strength of this research, therefore, is the use of qualitative research methods to provide important insights into the under researched area of early issues related primary surgical wounds relating to practical wound management and patient experience. We adopted a purposeful sampling strategy to ensure that perspectives were captured from a range of participants in relation to their primary surgical abdominal wound (10). Data produced from the interviews were supplemented by an analysis of existing RCT outcomes to ensure a comprehensive list of issues were initially generated, and therefore acted as a method of triangulation to increase the plausibility and dependability of the interview data (11).
However, it is important to note that these measures have only been pre-tested in relation to primary surgical wounds. Wounds that are intentionally left open or wounds that have developed problems are likely to require dressings that have advanced practical properties that are tailored to the wound requirements (15). Although participants were purposefully sampled, most had had a dressing of some kind (94%). A prospective real-time survey of dressings has demonstrated that this reflects current practice (16). In addition, these measures have only been pre-tested in relation to abdominal surgical wounds. However, characteristics of wound healing in this area are likely to be consistent with many other parts of the body.
Furthermore, these measures focus specifically on the experience of dressings -methods of wound closure (i.e. potentially leading to differential ease of removal of sutures or staples) may also affect patient experience, although this would require further investigation.
In summary, our measures can be completed both by patients and by health care professionals responsible for post-operative wound care. These measures will now be further developed to ensure they are appropriate and psychometrically-tested instruments, with a view to informing decision-making around dressings.

COMPETING INTERESTS
None declared.

DISCLAIMER
The views expressed are those of the authors and not necessarily those of the MRC, NHS, NIHR or the Department of Health.

Ethical approval for this work was granted by the Camden and King's Cross Research Ethics
Committee (14/LR/0640) on the 10th April 2014.

Wound Experience Questionnaire
We are interested in how your wound(s) have healed since your operation and your experience of having a dressing, as part of the Bluebelle study. Please complete this short questionnaire yourself. You can complete the questionnaire as soon as you feel ready, but ideally this will be within four days of having your operation. If there is more than one wound, please respond thinking about just one wound -either the main one or another wound if there have been any concerns about how it has been healing.
When you have completed the questionnaire, please return it in the pre-paid envelope provided.

To be completed by a healthcare professional up to 4 days after surgery Or To be completed by the participant up to 4 days after surgery if the participant is discharged before completion by a healthcare professional
If there is more than one wound, please respond thinking about just one wound -either the main one or another wound if there have been any concerns about how it has been healing. When you have completed the questionnaire, please return it in the pre-paid envelope provided.  The interpretation of each category is supported by illustrative quotes (See Table 2)

Data and findings consistent
Was there consistency between the data presented and the findings?
There is consistency between the data presented and the measures developed. The item tracking matrix provides an overview of the key findings and how these were used to develop the initial measure (See additional file) 31 Clarity of major themes Were major themes clearly presented in the findings?
The themes are clearly presented in the were regularly shared with the study team who suggested amendments to modify and reword items to improve understanding before further iterative testing with patients and health care professionals.
Results: Analyses of existing RCT outcomes and interviews produced a total of 69 issues.
Pre-testing and iterative revision established the need for two separate measures. One measure addresses health care professionals' experience of wound management in two key areas: exudate and its impact, and allergic reactions to the dressing. The other measure addresses patients' experience of wounds in seven key areas: wound comfort, dressing removal, dressings to protect the wound, impact on daily activities, ease of movement, anxiety about the wound and satisfaction with dressing. Each measure took less than five minutes to complete and were understood and acceptable to patients and health care professionals.
Conclusion: This in-depth study has developed two measures to assess practical management of primary surgical wounds and patient experience. Further work to test their validity and reliability and application to other settings is now required.

STRENGTHS AND LIMITATIONS OF THIS STUDY
• This is the first study to explore the important issues related to the practical management of primary surgical wounds and patient experience immediately following surgery.  1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59  • This study used robust methods to identify key issues of outcome that could be used to inform decision-making around dressings. Interviews provided a rich account of the key factors that affected wound management and patient experience while a purposeful sampling strategy ensured that perspectives were captured from a range of participants. Data produced from the interviews were supplemented by analysis of existing RCT outcomes to ensure a comprehensive list of issues was considered.
• Future work is needed to test the reliability, validity and sensitivity of the new measures.

BMJ Open
neither was there any benefit associated with particular dressings (3).
Decision-making around dressings may therefore need to be informed by other properties and qualities that dressings can offer, such as absorption of exudate, patient comfort, offering physical protection, facilitating wound observation, and meeting patients' desires for wound coverage (4). Whilst measures for assessment of wound cosmesis (in the longer term) are available (5,6), there is a lack of well-developed and validated measurement tools relating to practical wound management or patient experience (4,7,8). Such an instrument could be used to monitor the care of individual patients (e.g. assessing the ability of dressings to manage specific symptoms), audits (e.g. quality assurance) and research (e.g. comparing patient satisfaction).
The development of patient-reported outcome measures (PROMs) increasingly includes the use of qualitative research methods which provide the opportunity to elicit and characterise patients' experiences of their health conditions and treatment (9,10). Qualitative methods can also define health professionals' experiences of care and management (11). Data can be supplemented by expert input and studies in published literature (12,13). This article describes the development of measures to assess practical wound management issues, symptoms and patient experience associated with primary surgical wounds.

Study design
Measures were developed according an existing framework for developing PROMs (14,15), also incorporating guidance on eliciting health domain concepts using qualitative methodologies (12,13,16). The study is reported according to qualitative reporting guidelines (See Supplementary File 1). Phase 1 aimed to produce a comprehensive list of potential issues relating to wound and dressing experience and practical management issues.  1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59  Interviews were audio-recorded and transcribed in full. Transcripts were imported into NVivo (version 10). All data relating to outcomes and issues of importance to patients that were relevant to dressing use and the practical management of the wound in the initial time period post-surgery were assigned labels (coded) by two experienced qualitative researchers.
Data were analysed using techniques of constant comparison derived from grounded theory methodology, and emerging codes across the dataset were then compared to look for shared or disparate views among participants (18). A subset of approximately half of the interviews (n = 19) was double coded by third experienced researcher to highlight any differences in the interpretation of codes (12). Data collection and analysis continued until the team were confident that saturation had been reached i.e. no more patterns or themes emerged from the data (19).

Extraction of information from three systematic reviews
Systematic reviews were purposefully selected to identify RCTs measuring outcomes relating to patient experience and management of wound healing. Since the wider Bluebelle study was exploring whether a trial comparing different types of dressings (including dressings versus the novel use of tissue glue as a dressing) to no dressing was possible, we selected three recent systematic reviews (4,20,21) to identify RCTs which included outcomes relevant to both dressings and the use of tissue adhesive. Although not published at the time of conducting this work, additional references from an updated version of one of the systematic reviews (7) were also provided. Published papers reporting the RCTs included in the systematic reviews were obtained where possible. Relevant data from the RCT reports were then extracted on the outcome (as described by the authors), the verbatim wording to measure outcome, who reported the outcome, the measurement scale and the assessment time point. Attempts were made to contact the authors for more information.

Synthesis of findings from interviews and data extraction
A list of issues from the analysis of the interviews and analysis of existing RCT outcomes was collated into an item tracking matrix, in line with guidance for developing PROMs (22). This is available in the online appendix (see Supplementary File 3). The study team agreed on a set of words or phrases to reflect each issue and also noted additional phrasing made by participants in a subsequent column (12). Issues which were conceptually similar were organised into categories. For instance, issues such as 'itchiness/irritation', 'presence of pulling sensation', and 'tightness of wound' were mapped into a 'wound comfort' category. and age, gender, ethnicity, type of surgery, dressing type and location (for patients).
Cognitive interviews are used widely in questionnaire development (12) and involve asking respondents to verbalise their thoughts while answering questions (24). This methodology enabled us to explore the acceptability of the measure and coverage of patients and health care professionals' concerns (in terms of language, accuracy, and relevance, as well as layout (13)). During each interview, participants were asked to complete the measure by reading each item aloud and commenting on their understanding. Interviews were guided by a series of probes (e.g. 'What does this item mean to you?', 'Are there other ways you would describe it?'; (24)). Participants' body language (such as nodding or frowning) was also observed and prompted further discussion about specific items (12). A copy of the topic guide is available (See Supplementary File 2).
The qualitative team maintained detailed field notes from each interview, describing suggestions for modifications and improvements to the provisional measures.
Operationalisation and modification of the measures was an iterative process. Findings from the interviews and suggestions for amendments were regularly disseminated to the Bluebelle Study Group, which consisted of a multidisciplinary group of health care professionals, including surgeons, health services researchers and research nurses. Each stage of feedback informed amendments to modify and reword items to improve understanding, which was repeated following efforts to revise questions and eliminate problems (24). This process continued until no new issues were identified and no further refinements were believed to be necessary.

Interviews
A total of 39 interviews were conducted between July 2014 and July 2015. Interviews were conducted in person (n=10), unless patients preferred to be interviewed via telephone  Table 1.

Extraction of information from three systematic reviews
Published papers for 26 studies that included outcomes relating to patient experience and management of wound healing were identified from the three systematic reviews .
Only two studies included a validated instrument, or modification of a validated instrument, to assess outcomes (27,41). These were for long term scarring and cosmesis (5,6).
However, no studies reported using validated measures relating to issues associated with practical wound management and patient experiences in the early post-operative period.

Synthesis of findings from interviews and data extraction
When describing experiences in the interviews, patients commented on several factors that affected perceptions of how well their wound was healing, including how it felt (tightness, pain, and itchiness) and whether any fluid had leaked from the wound. Analysis of existing RCT outcomes showed these issues had been captured in some previous (unvalidated) outcomes.
All patients had at least one dressing applied after surgery, although this varied between adhesive coverings (absorptive or non-absorptive) and tissue adhesive as a dressing. Both the interviews and the analysis of existing RCT outcomes highlighted that multiple practical advantages of dressing use (including ability to contain exudate and ease of removal). The interviews also demonstrated that there were psychological factors that affected dressing experience and satisfaction (i.e. anxiety about cleanliness of the wound). Patients with tissue adhesive as a dressing commented that they had been surprised that their wounds had been dressed this way (rather than adhesive dressings which they had had in the past for other wounds). However, these patients stated that compared to past experiences of adhesive dressings, they liked how glue was transparent, waterproof, did not require multiple applications and came off naturally.
The interviews and the analysis of existing RCT outcomes produced a total of 69 issues.
These were grouped into ten broad categories: wound comfort, exudate and its impact, allergic reactions to the dressing, dressing removal, dressings to protect the wound, impact on daily activities, ease of movement, anxiety about the wound, satisfaction with dressing and wound appearance. Table 2 provides illustrative quotes for the categories identified.   Table 1.
Interviews highlighted issues with content in the initial measure. For example, items regarding the colour of the wound exudate were removed. Questions were rephrased to focus on the experience of having a dressing rather than general recovery after surgery (i.  Overall, the final versions of the measures were well received. In addition, 96% of participants stated that each measure took less than five minutes to complete.

DISCUSSION
This paper describes the development of two measures for assessing wound management and experience. The Wound Management Questionnaire assesses practical issues early after surgery for completion by health care professionals and The Wound Experience Questionnaire assesses symptoms and patient satisfaction with their wound and dressings.
These measures were developed using a mixed methods approach, including data extraction from 26 published RCTs and interviews with 64 patients and 15 health care professionals.
Final versions of the measures were easily completed and acceptable to patients and health Given the absence of evidence supporting the effectiveness of dressings for the prevention of SSIs, decision-making around dressings needs to be informed by issues such as managing wound exudate, offering physical protection and meeting patients' desires for wound coverage. However, systematic reviews have highlighted a lack of meaningful outcome data on wound symptom management and patient experiences of primary surgical wounds and acceptability of dressings (4,7,8). To our knowledge, this is the first study to explore these important issues in patients with closed primary surgical wounds. The measures are intended to be used in future studies including a wide variety of primary abdominal wounds such as those created during elective or acute surgery, surgery for benign or malignant disease and bowel resection of obstetric procedures. Such studies will always record the patient group which will be important to consider when looking at the results of the measures.
True patient-centred outcome measures require full consideration of patients' experiences and views' (9,10). The main strength of this research, therefore, is the use of qualitative research methods to provide important insights into the under researched area of early issues related primary surgical wounds relating to practical wound management and patient experience. We adopted a purposeful sampling strategy to ensure that perspectives were captured from a range of participants in relation to their primary surgical abdominal wound (12). Data produced from the interviews were supplemented by an analysis of existing RCT outcomes to ensure a comprehensive list of issues were initially generated, and therefore acted as a method of triangulation to increase the plausibility and dependability of the interview data (13). (i.e. potentially leading to differential ease of removal of sutures or staples) may also affect patient experience, although this would require further investigation.
In summary, our measures can be completed both by patients and by health care professionals responsible for post-operative wound care. These measures will now be further developed to ensure they are appropriate and psychometrically-tested instruments, with a view to informing decision-making around dressings.

ACKNOWLEDGEMENTS
We would like to thank all participants for giving up their time to take part in this study, and members of the Bluebelle study team for their helpful comments on the earlier drafts of the measures.

Wound Experience Questionnaire
We are interested in how your wound(s) have healed since your operation and your experience of having a dressing, as part of the Bluebelle study. Please complete this short questionnaire yourself. You can complete the questionnaire as soon as you feel ready, but ideally this will be within four days of having your operation. If there is more than one wound, please respond thinking about just one wound -either the main one or another wound if there have been any concerns about how it has been healing.
When you have completed the questionnaire, please return it in the pre-paid envelope provided.

Wound Management Questionnaire
To be completed by a healthcare professional up to 4 days after surgery Or To be completed by the participant up to 4 days after surgery if the participant is discharged before completion by a healthcare professional If there is more than one wound, please respond thinking about just one wound -either the main one or another wound if there have been any concerns about how it has been healing. When you have completed the questionnaire, please return it in the pre-paid envelope provided.  1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59   Derivation of themes Were themes identified in advance or derived from the data?
Issues which were conceptually similar were organised into categories. For instance, issues such as 'itchiness/irritation', 'presence of pulling sensation', and 'tightness of wound' were mapped into a 'wound comfort' category.
(See Table 2)  27 Software What software, if applicable, was used to manage the data?
NVivo (version 10) was used to analyse the data (See page 6) 28 Participant checking Did participants provide feedback on the findings? Full results were not sent out to all participants to gain respondent validation.

29
Quotations presented Were participant quotations presented to illustrate the themes / findings? Was each quotation identified? e.g. participant number The interpretation of each category is supported by illustrative quotes (See Table 2)

Data and findings consistent
Was there consistency between the data presented and the findings?
There is consistency between the data presented and the measures developed. The item tracking matrix provides an overview of the key findings and how these were used to develop the initial measure (See additional file) 31 Clarity of major themes Were major themes clearly presented in the findings?