- Email: [email protected]
Impact of the PROCESS guideline on the reporting of surgical case series: A before and after study
International Journal of Surgery 45 (2017) 92e97
Contents lists available at ScienceDirect
International Journal of Surgery journal homepage: www.journal-surgery.net
Before and After Study
Impact of the PROCESS guideline on the reporting of surgical case series: A before and after study R.A. Agha a, M.R. Borrelli a, *, R. Farwana b, T. Kusu-Orkar c, M.C. Millip d, R. Thavayogan e, J. Garner f, N. Darhouse a, D.P. Orgill g a
Department of Plastic Surgery, Guy's and St. Thomas' NHS Foundation Trust, London, UK School of Medicine and Dentistry, University of Birmingham, Birmingham, UK School of Medicine and Dentistry, University of Liverpool, Liverpool, UK d Bart's and the London School of Medicine and Dentistry, Queen Mary and Westfield University, London, UK e School of Medicine, University of Nottingham, Nottingham, UK f School of Medicine and Dentistry, University of Bristol, Bristol, UK g Division of Plastic Surgery, Brigham and Women's Hospital, Boston, MA, USA b c
h i g h l i g h t s This is the first assessment of the PROCESS guidelines since its publication. There was a 5% increase in concordance with the PROCESS guidelines after publication. Research is needed to identify and navigate existing barriers to greater compliance. Authors, reviewers and editors should adhere to the guidelines to boost reporting quality.
a r t i c l e i n f o
a b s t r a c t
Article history: Received 20 July 2017 Accepted 21 July 2017 Available online 29 July 2017
Introduction: The PROCESS guideline was developed in 2016 through expert Delphi consensus. It aimed to improve the quality of reporting of surgical case series. This study assessed the impact of the introduction of the PROCESS guideline on reporting for surgical case series submitted to three journals. Methods: 20 case series published in the International Journal of Surgery Case Reports (IJSCR), the International Journal of Surgery (IJS) or the Annals of Medicine and Surgery (AMS) in September to December 2016, prior to the introduction of the PROCESS guideline (the pre-PROCESS period), were randomly identified and scored against the PROCESS criteria. Two authors independently scored each article a total score out of 29, the 'PROCESS score' (expressed as a percentage). Scores for the two researchers were compared and consensus was reached to achieve a final score set. The process was repeated for the January 2017 to April 2017 issues of the three journals, post PROCESS implementation (the post-PROCESS period). Results: The mean PROCESS score was 80% (range 66e90%) for the pre-PROCESS period and 84% (range 72e95%) for the post-PROCESS period, a 4% relative increase [STATS]. The Cohen's Kappa score between researchers was 0.907 implying very substantial agreement. Conclusion: Implementation of the PROCESS guideline resulted in a 5% improvement in the reporting quality of surgical case series published in three journals. Further research is needed to identify and successfully navigate existing barriers to greater compliance. Authors, reviewers and editors should adhere to the guidelines to boost reporting quality. Journals should develop their policies and guide for authors to incorporate the guideline and mandate compliance. © 2017 IJS Publishing Group Ltd. Published by Elsevier Ltd. All rights reserved.
Keywords: PROCESS Reporting criteria Case series
1. Introduction
* Corresponding author. E-mail address: [email protected] (M.R. Borrelli).
The surgical case series is a traditional stalwart of the literature. It has been defined as: “an observational study of an uncontrolled group of patients collected or sampled on the basis of a specific disease/
http://dx.doi.org/10.1016/j.ijsu.2017.07.079 1743-9191/© 2017 IJS Publishing Group Ltd. Published by Elsevier Ltd. All rights reserved.
R.A. Agha et al. / International Journal of Surgery 45 (2017) 92e97
exposure/intervention or a specific outcome of interest”. [1] Despite their designation as level four evidence in drive for evidence-based medicine (EBM) over the last few decades, they are still highly prevalent and have a role to play [1]. Recent work by our group found that 25e40% of the research output in 2013 of five surgical specialties published in their top three journals (by impact factor) was a case series [2]. Surgical case series provide important insights and initial evidence on new diseases, the mechanisms of disease and their natural history, rare diseases and rare manifestations and their investigation and management, side effects, drug reactions, the management of complications and reminders of important lessons. They are important in emergency scenarios or with population sub-sets where trials cannot occur e.g. children and those with dementia. Early and initial evidence for new treatments and technologies usually comes in the form of a case series, for example vacuum assisted closure therapy [3] or use of propopanol for haemangiomas [4]. Case series have even been used as a means of putting forth a new classification system, such as Mulliken and Glowacki's work with haemangiomas and vascular malformations [5]. Case series are cheap, economical and easier to conduct than other observational studies such as cohort studies, or randomised controlled trials. Indeed, they may act as a ‘launch pad’ into those types of studies. This is recognised by the IDEAL framework where a form of high quality prospective case series is the basis by which a new technique achieves technical stability prior to multicentre take-up and refinement of its indications. However, case series do suffer with inherent bias and recent systematic reviews have found reporting quality and transparency issues [6]. Our group recently completed a systematic review on the reporting quality of case series in surgery over the period 1990e2014 [7]. From 92 articles that met the inclusion criteria, methodological and reporting issues identified were: failure to use standardised definitions (57%), missing or selective data (66%), poor transparency with incomplete reporting (70%), failure to mention whether alternate study designs were considered (11%) and ‘other’ issues (52%) including the failure to clearly define the patient population under investigation, selection bias insufficient followup time, need for validated outcomes. Readers need complete, clear and transparent information. Authors may benefit from guidelines outlining the minimum criteria necessary to report in a case series to adequate and effectively report their case series. We published the PROCESS guideline in December 2016 to aid the writing of surgical case series [8]. It was developed through a Delphi consensus exercise amongst a 59strong international expert group. At the time of writing, it has been cited 40 times according to Google Scholar, has been downloaded 1286 times, is listed on Equator network website (www. equator-network.org) and has its own dedicated website www. processguideline.com [9]. The impact of the PROCESS guidelines has not been previously assessed. It is important to evaluate the effect PROCESS guidelines have had on reporting case series to assess their utility as a tool that can be used by journal editors and reviewers as guidance, as well as guidance to authors. This study is the first study to assesses the impact of the introduction of the PROCESS guideline on reporting of surgical case series submitted to three international journals. 2. Methods 2.1. Article selection 20 case series published in the International Journal of Surgery Case Reports (IJSCR), the International Journal of Surgery (IJS) or the Annals of Medicine and Surgery (AMS) were assessed in the period
93
September to December 2016. This was prior to the introduction of the PROCESS guideline (the pre-PROCESS period). The articles were randomly identified and scored against the PROCESS criteria, by searching the journal data bases using the key word ‘case series’ in the specified date range. This generated a list of 219 potential articles across the three articles, numbered sequentially for randomisation (N ¼ 69 in IJS, N ¼ 32 in AMS, N ¼ 43 in IJSCR). A computer generated random number between 1 and 219 was used to select the article for inclusion. The article was included provided it was a case series. Articles after the introduction of the PROCESS guideline were only included if they reported adherence to the PROCESS criteria. These three journals were chosen as they were the only ones to have implemented the PROCESS guidelines at that time. 2.2. Article coding Each case series was scored against the 29 items in the PROCESS criteria. An item was given 1 point if it satisfied all criteria in the item, 0.5 if there was partial completion, 0 e for not addressing the item at all and NA (not applicable) if this item did not apply. Each item could score a maximum theoretical ‘PROCESS score’ of 29 (assuming all criteria are applicable). Points would be deducted from this maximum in the denominator if they were deemed not relevant to that specific case series. Table 1 below shows how the total score out of 29 would be reached for the PROCESS criteria: Two independent teams performed scoring. Scores for the two teams were compared and consensus was reached to achieve a final score set. The methodology was revised to involve just two authors performing all scoring. Before scoring the authors discussed each PROCESS point in detail, scored a pilot case series independently and discussed discrepancies to resolve misunderstandings, before independent scoring on the 40 articles was carried out. All scoring was conducted on a pre-formatted excel database. Once scoring was completed ‘PROCESS score’ for each article was calculated and expressed as a percentage. The percentage of each of the 29 PROCESS items answered correctly out of the theoretical maximum of 40 was also calculated. Prior to starting work, data collectors were given a briefing and a standardised data extraction template. Training was provided to the lead data collectors (MB and RF). 2.3. Statistics SPSS was used for statistical analysis (version 23, SPSS Inc, Chicago, IL). The Shapiro Wilks test was used to test for normally distributed data. Although medians and ranges are usually appropriate for descriptive data, means and standard deviations were given for normally distributed data. For normally distributed data the independent-samples t-test was used to compare pre- and post scores (two-tailed). Cohen's Kappa was used to assess agreement between researchers. A score of greater than 0.5 was considered moderate agreement. 3. Results The mean PROCESS score was 79.9% (standard deviation ± 6.9, range 66e90) for the pre-PROCESS period and 84.1% (standard deviation ± 6.2, range 72e95) for the post-PROCESS period, a 5% relative increase. Tests for normality confirmed a normal or Gaussian distribution and this increase was found to be statistically significant (t(19) ¼ 2.164, P ¼ 0.043. This data is summarised in Table 2: The Cohen's Kappa score was 0.907, implying very substantial agreement. Fig. 1 below shows how compliance with the different items changed before and after PROCESS:
94
R.A. Agha et al. / International Journal of Surgery 45 (2017) 92e97
Table 1 The PROCESS guideline checklist and how marks were awarded. Topic
Item
Checklist item description
Title
1
Abstract
2a 2b
The words “case series” and the area of focus should appear in the title (e.g. disease, exposure/intervention or outcome). Introduction - what is the unifying theme of the case series. Methods - describe what was done, how and when was it done and by whom. Results - what was found. Conclusion - what have we learned and what does it mean
2c 2d Introduction
3
Methods
4a
4b
4c
4d
4e
4f
4g
4h
4i
4j
Results
5a
5b
5c
5d
5e
Discussion
6a
How marks were awarded
0.5 points for case series 0.5 points for area of focus 1 point if theme mentioned, else 0 0.5 points for ‘what’ 0.5 points for ‘how’ 1 point for results found, else 0 1 point if mention what learnt, 0 points if not Explain the scientific background and rationale for the case series. What is 0.5 points for referral to literature the unifying theme - common disease, exposure, intervention and outcome, 0.5 points for study need etc. Why is this study needed? Registration and ethics - state the research registry number in accordance 0.5 e for mentioning ethics 0.5 e for registering study with the declaration of Helsinki - “Every research study involving human subjects must be registered in a publicly accessible database” (this can be obtained from; ResearchRegistry.com or ClinicalTrials.gov or ISRCTN). State whether ethical approval was needed and if so, what the relevant judgement reference was? Study design - state the study is a case series and whether prospective or 0.25 - case series retrospective in design, whether single or multi-centre and whether cases 0.25 e if mention prospective/retrospective 0.25 e for mentioning centre are consecutive or non-consecutive. 0.25 e for mentioning whether or not consecutive 1/3 e for mentioning setting where recruit the patients Setting - describe the setting(s)and nature of the institution in which the 1/3 points for mentioning when recruited patient was managed; academic, community or private practice setting? Location(s), and relevant dates, including periods of recruitment, exposure, 1/3 points for mentioning follow-up follow-up, and data collection 1/3 e points for mentioning inclusion/exclusion criteria Participants - describe the relevant characteristics of the participants 1/3 points for mentioning smoking or social health (comorbidities, tumour staging, smoking status, etc). State any eligibility 1/3 points for mentioning comorbidities (inclusion/exclusion) criteria and the sources and methods of selection of participants. Describe length and methods of follow-up. Pre-intervention considerations e.g. Patient optimisation: measures taken 1 point if mention, if applicable, else 0 prior to surgery or other intervention e.g. treating hypothermia/ hypovolaemia/hypotension in burns patients, ICU care for sepsis, dealing with anticoagulation/other medications and so on. Types of intervention(s) deployed and reasoning behind treatment offered 1/3 points for intervention 1/3 points for reasoning (pharmacological, surgical, physiotherapy, psychological, preventive) and 1/3 points for concurrent treatment concurrent treatments (antibiotics, analgesia, anti-emetics, nil by mouth, VTE prophylaxis, etc). Medical devices should have manufacturer and model specifically mentioned. 0.5 points for what was done specifically Peri-intervention considerations - administration of intervention (what, 0.5 points for how this was done where, when and how was it done, including for surgery; anaesthesia, patient position, use of tourniquet and other relevant equipment, preparation used, sutures, devices, surgical stage (1 or 2 stage, etc). Pharmacological therapies should include formulation, dosage, strength, route and duration). 1 point for mentioning who performed the procedure, else 0 Who performed the procedures - operator experience (position on the learning curve for the technique if established, specialisation and prior relevant training). Quality control - what measures were taken to reduce inter or intra1 point for mentioning effort to reduce inter or intraoperator variation. What measures were taken to ensure quality and operator variation consistency in the delivery of the intervention e.g. independent observers, lymph node counts, etc Post-intervention considerations e.g. post-operative instructions and place 0.5 points for mentioning a post-operative care package 0.5 points for mentioning location of follow-up of care. Important follow-up measures - diagnostic and other test results. Future surveillance requirements - e.g. imaging surveillance of endovascular aneurysm repair (EVAR) or clinical exam/ultrasound of regional lymph nodes for skin cancer. Participants - reports numbers involved and their characteristics 1/3 points e for mentioning participant number (comorbidities, tumour staging, smoking status, etc). 1/3 points for mentioning smoking status 1/3 points for mentioning comorbidities Any changes in the interventions during the course of the case series (how 0.5 points for change mentioned if applicable, 0.5 points for why change has it evolved, been tinkered with, what learning occurred, etc) together with rationale and a diagram if appropriate. Degree of novelty for a surgical else NA technique/device should be mentioned and a comment on learning curves should be made for new techniques/devices. 1 point for mentioning follow-up time, else 0 Outcomes and follow-up - Clinician assessed and patient-reported outcomes (when appropriate) should be stated with inclusion of the time periods at which assessed. Relevant photographs/radiological images should be provided e.g. 12 month follow-up. Where relevant - intervention adherence/compliance and tolerability (how 0.5 if lost patients to follow-up was this assessed). Describe loss to follow-up (express as a percentage) and 0.5 points if mention why or, 1 point if no lost to follow up any explanations for it. Complications and adverse or unanticipated events. Described in detail and 0.5 point for mentioning complications 0.5 points for mentioning how managed, ideally categorised in accordance with the Clavien-Dindo Classification. How they were prevented, diagnosed and managed. Blood loss, operative NA if not applicable time, wound complications, re-exploration/revision surgery, 30-day postop and long-term morbidity/mortality may need to be specified. Summarise key results 1 point for summarising results, else 0
R.A. Agha et al. / International Journal of Surgery 45 (2017) 92e97
95
Table 1 (continued ) Topic
Item
Checklist item description
6b
1 point for discussing results with previous literature, else 0 Discussion of the relevant literature, implications for clinical practice guidelines, how have the indications for a new technique/device been refined and how do outcomes compare with established therapies and the prevailing gold standard should one exist and any relevant hypothesis generation. Strengths and limitations of the study 0.5 points for strengths 0.5 points for limitations The rationale for any conclusions? 1 point for rational, else 0 State the key conclusions from the study 1 point for key conclusions, else 0 State what needs to be done next, further research with what study design. 0.5 points for mentioning need for future work, 0.5 points if study design mentioned State any conflicts of interest 1 point if mention, else 0 State any sources of funding 1 point if mention Total ¼ 29
6c
Conclusions
6d 7a 7b
Additional Information
8a 8b
How marks were awarded
Table 2 Summary results for the pre and post PROCESS periods. Period
Mean
S.D.
t
df
P value
Pre-PROCESS (n ¼ 20) Post-PROCESS (n ¼ 20)
79.9 84.1
6.9 6.2
2.053
38
0.047
Abbreviations: SD e standard deviation, S.D. e standard deviation.
From Fig. 1 we can see that the areas that improved the most and these are shown in Table 3 below. Table 4 below shows the top five areas of poorest compliance in the pre and post PROCESS periods. It can be noted that items 4e, 7b and 4h are in the top three reasons for both groups. 4. Discussion Our study showed a 5% relative increase in the compliance with the PROCESS guideline before and after its implementation. In both periods the top three items most poorly complied with were similar (Table 4). The reasons for this consistency is uncertain and was not of specific focus to the research question posed here. In both pre- and post-PROCESS periods point 4h ‘who performed the procedure and their level of experience’ is often missing
from published case series. Authors may feel this is irrelevant, especially for less complex procedures which may be performed by multiple surgeons with different experience levels and credentials. Also, if there was no significant pre-intervention consideration, authors may simply not mention it. The focus of the PROCESS guidelines is to guide authors in how to transparently report both what has and has not been done. The need to report a negative in this context may be underappreciated by authors. The operator and their level of experience, however, can be a strong factor determining surgical outcomes and is therefore important this information is portrayed to readers. Another area poorly reported in case series is point 7b ‘what needs to be done and any future work’. Authors may feel their case series is a definitive guide to practice and hence there isn't a need for further work. However, authors may simply be omitting this information for lack of paying close enough attention to the guidelines. This data is useful going forward as it will provide a focus for future efforts to raise compliance, especially if peerreviewers and editors are aware of these consistent issues. The responsibility for adequate and transparent reporting is not only that of authors but also of the editors and reviewers of journals. The number of reporting guidelines has increased substantially in recent years. The Equator network mentions 362 at the
Fig. 1. Compliance with the 8 PROCESS items before and after the guideline's introduction.
96
R.A. Agha et al. / International Journal of Surgery 45 (2017) 92e97
Table 3 The most items that improved the most following introduction of the PROCESS guideline. Item
Item description
Improvement in compliance
4e
Pre-intervention considerations e.g. Patient optimisation: measures taken prior to surgery or other intervention e.g. treating hypothermia/hypovolaemia/hypotension in burns patients, ICU care for sepsis, dealing with anticoagulation/ other medications and so on. Who performed the procedures - operator experience (position on the learning curve for the technique if established, specialisation and prior relevant training). Study design - state the study is a case series and whether prospective or retrospective in design, whether single or multi-centre and whether cases are consecutive or non-consecutive. The words “case series” and the area of focus should appear in the title (e.g. disease, exposure/intervention or outcome). Registration and ethics - state the research registry number in accordance with the declaration of Helsinki - “Every research study involving human subjects must be registered in a publicly accessible database” (this can be obtained from; ResearchRegistry.com or ClinicalTrials.gov or ISRCTN). State whether ethical approval was needed and if so, what the relevant judgement reference was?
253%
4h 4b 1 4a
135% 51% 47% 42%
Table 4 The top five areas of poorest compliance in the pre and post SCARE periods. Pre-PROCESS (criterion)
Post-PROCESS (criterion)
1
17% - Pre-intervention considerations (item 4e).
2
20% - who performed the procedure (item 4h).
3
35% - State what needs to be done next, further research with what study design (item 7b)
4
55% - Study design - state the study is a case series and whether prospective or retrospective in design, whether single or multi-centre and whether cases are consecutive or non-consecutive. (item 4B)
5
55% - Participants - describe the relevant characteristics of the participants (comorbidities, tumour staging, smoking status, etc). State any eligibility (inclusion/exclusion) criteria and the sources and methods of selection of participants. Describe length and methods of follow-up (item 4D)
40% - State what needs to be done next, further research with what study design (item 7b). 47% - Who performed the procedures - operator experience (position on the learning curve for the technique if established, specialisation and prior relevant training (item 4h). 60% - Pre-intervention considerations e.g. Patient optimisation: measures taken prior to surgery or other intervention e.g. treating hypothermia/ hypovolaemia/hypotension in burns patients, ICU care for sepsis, dealing with anticoagulation/other medications and so on (item 4e) 63% - Setting - describe the setting(s)and nature of the institution in which the patient was managed; academic, community or private practice setting? Location(s), and relevant dates, including periods of recruitment, exposure, follow-up, and data collection (item 4c) 67% - Participants - reports numbers involved and their characteristics (comorbidities, tumour staging, smoking status, etc) (item 5a)
time of writing (July 2017) [10]. Surgical journals have been slow to take up reporting guidelines. Previous research has shown how the majority (62%) of the 193 surgical journals listed in the surgery category in the Journal Citation Report 2014, made no mention of reporting guidelines within their guide for authors at all [11]. Evidence on the effectiveness of reporting guidelines is mixed. A Cochrane systematic review focusing on randomised controlled trials and CONSORT was conducted by Turner et al. [12] It found that 25 of 27 outcomes assessing completeness of reporting favoured CONSORT-endorsing journals over non-endorsers, of which five were statistically significant. Simply endorsing a guideline doesn't mean that 100% compliance will be achieved and our data support this Cochrane review. In IJS, AMS and IJSCR all authors are required to submit a completed checklist for case series as well as other common manuscript types and this is available to peerreviewers and the editor. Authors are unable to submit a manuscript until they have met this requirement. The magnitude of effect (i.e. a 5% increase) is similar to what we have seen for other guidelines where submission of a completed checklist was made mandatory as part of a wider policy initiative [13]. Further research is needed to understand how we can increase compliance yet further from the 84% achieved here. Solutions may be more forthcoming once more journals adopt the guideline and assess their own data on compliance. There are several limitations of our study. Firstly, these analyses were conducted soon after the guidelines were published in December 2016 and hence awareness will be more limited and authors, reviewers and editors will still be getting used to them. A
follow-up study further down the line is planned which will be a useful additional comparison. The analysis is comparing only one pre-PROCESS and one post-PROCESS period, so chance and Hawthorne effects can play a role. The journals chosen are only three of the many journals in which case series are published, so generalisability and external validity is limited somewhat. Reviewers were not blinded to the articles (authors would mention their work was compliant with the PROCESS criteria in the text so this was not feasible) and articles were not reviewed in randomised order (i.e. pre-PROCESS first and then post-PROCESS) and there is a degree of subjectivity in marking.
5. Conclusion Implementation of the PROCESS guideline resulted in a 5% improvement in the reporting quality of surgical case series published in three journals. Authors, reviewers and editors should adhere to the guidelines to boost reporting quality and journals should develop their policies, submission processes and guide for authors to incorporate the guideline. Other journals publishing surgical case reports should consider utilising the PROCESS guideline and how they can achieve compliance with it during the process of submission and peer-review.
Ethical approval Not required for this before after study.
R.A. Agha et al. / International Journal of Surgery 45 (2017) 92e97
Sources of funding None. Author contribution RAA developed the concept, analysed and interpreted the data, wrote the first draft and revised it in line with feedback from other authors. TKO, MMC, RF, MRB, RT and JG were involved in data collection. MRB conducted the statistical analyses. ND and DO was involved in data interpretation. All authors were involved in critically reviewing the manuscript draft and approving the final version for submission. Conflicts of interest None. Research registration unique identifying number (UIN) This is not a human study. Guarantor RAA. References [1] Oxford Centre for Evidence-based Medicine e Levels of Evidence, March 2009 [online]. Available at, http://www.cebm.net/oxford-centre-evidence-basedmedicine-levels-evidence-march-2009/. Accessed 16 July 2017.
97
[2] R.A. Agha, M. Devesa, K. Whitehurst, A.J. Fowler, R. Coe, G. Wellstead, D.P. Orgill, P. McCulloch, Levels of evidence in plastic surgery - bibliometric trends and comparison with five other surgical specialties, Eur. J. Plastic Surg. 39 (5) (2016) 365e370. [3] L.C. Argenta 1, M.J. Morykwas, Vacuum-assisted closure: a new method for wound control and treatment: clinical experience, Ann. Plast. Surg. 38 (6) (1997 Jun) 563e576. aute -Labre ze, E. Dumas de la Roque, T. Hubiche, F. Boralevi, J.B. Thambo, [4] C. Le A. Taïeb, Propranolol for severe hemangiomas of infancy, N. Engl. J. Med. 358 (24) (2008 Jun 12) 2649e2651, http://dx.doi.org/10.1056/NEJMc0708819. [5] J.B. Mulliken, J. Glowacki, Hemangiomas and vascular malformations in infants and children: a classification based on endothelial characteristics, Plast. Reconstr. Surg. 69 (3) (1982 Mar) 412e422. [6] R.A. Agha, A.J. Fowler, C. Herlin, T. Goodacre, D.P. Orgill, Use of autologous fat grafting for breast reconstruction: a systematic review and meta-analysis, J. Plastic Reconstr. Surg. 68 (2015) 143e161. [7] R. Agha, A. Fowler, S. Lee, B. Gundogan, B. Whitehurst, H. Sagoo, K.J.L. Jeong, D. Altman, D. Orgill, A systematic review of the methodological and reporting quality of care series in surgery, Br. J. Surg. 103 (10) (2016) 1253e1258. [8] R.A. Agha, A.J. Fowler, S. Rajmohan, I. Barai, D.P. Orgill, PROCESS Group, Preferred reporting of case series in surgery; the PROCESS guidelines, Int. J. Surg. 36 (Pt A) (2016) 319e323. [9] Google Scholar: The SCARE statement: consensus-based surgical case report guidelines [online]. Available at: https://scholar.google.com/citations?view_ op¼view_citation&hl¼en&user¼Ig4lUDEAAAAJ&citation_for_ view¼Ig4lUDEAAAAJ:WA5NYHcadZ8C (Accessed 16 July 2017). [10] Equator network [online]. Available at: http://www.equator-network.org/. [11] R.A. Agha, I. Barai, S. Rajmohan, S. Lee, M.O. Anwar, A.J. Fowler, D.P. Orgill, D.G. Altman, Support for reporting guidelines in surgical journals needs improvement: a systematic review, Int. J. Surg. (2017) article in press. [12] L. Turner, L. Shamseer, D.G. Altman, L. Weeks, J. Peters, T. Kober, et al., Consolidated standards of reporting trials (CONSORT) and the completeness of reporting of randomised controlled trials (RCTs) published in medical journals, Cochrane database Syst. Rev. 11 (2012 Jan). MR000030. [13] R. Agha, A. Fowler, C. Limb, K. Whitehurst, R. Coe, H. Sagoo, D. Jafree, C. Chandrakumar, B. Gundogan, Impact of the mandatory implementation of reporting guidelines on reporting quality in a surgical journal: a before and after study, Int. J. Surg. 30 (2016) 169e172.
Contents lists available at ScienceDirect
International Journal of Surgery journal homepage: www.journal-surgery.net
Before and After Study
Impact of the PROCESS guideline on the reporting of surgical case series: A before and after study R.A. Agha a, M.R. Borrelli a, *, R. Farwana b, T. Kusu-Orkar c, M.C. Millip d, R. Thavayogan e, J. Garner f, N. Darhouse a, D.P. Orgill g a
Department of Plastic Surgery, Guy's and St. Thomas' NHS Foundation Trust, London, UK School of Medicine and Dentistry, University of Birmingham, Birmingham, UK School of Medicine and Dentistry, University of Liverpool, Liverpool, UK d Bart's and the London School of Medicine and Dentistry, Queen Mary and Westfield University, London, UK e School of Medicine, University of Nottingham, Nottingham, UK f School of Medicine and Dentistry, University of Bristol, Bristol, UK g Division of Plastic Surgery, Brigham and Women's Hospital, Boston, MA, USA b c
h i g h l i g h t s This is the first assessment of the PROCESS guidelines since its publication. There was a 5% increase in concordance with the PROCESS guidelines after publication. Research is needed to identify and navigate existing barriers to greater compliance. Authors, reviewers and editors should adhere to the guidelines to boost reporting quality.
a r t i c l e i n f o
a b s t r a c t
Article history: Received 20 July 2017 Accepted 21 July 2017 Available online 29 July 2017
Introduction: The PROCESS guideline was developed in 2016 through expert Delphi consensus. It aimed to improve the quality of reporting of surgical case series. This study assessed the impact of the introduction of the PROCESS guideline on reporting for surgical case series submitted to three journals. Methods: 20 case series published in the International Journal of Surgery Case Reports (IJSCR), the International Journal of Surgery (IJS) or the Annals of Medicine and Surgery (AMS) in September to December 2016, prior to the introduction of the PROCESS guideline (the pre-PROCESS period), were randomly identified and scored against the PROCESS criteria. Two authors independently scored each article a total score out of 29, the 'PROCESS score' (expressed as a percentage). Scores for the two researchers were compared and consensus was reached to achieve a final score set. The process was repeated for the January 2017 to April 2017 issues of the three journals, post PROCESS implementation (the post-PROCESS period). Results: The mean PROCESS score was 80% (range 66e90%) for the pre-PROCESS period and 84% (range 72e95%) for the post-PROCESS period, a 4% relative increase [STATS]. The Cohen's Kappa score between researchers was 0.907 implying very substantial agreement. Conclusion: Implementation of the PROCESS guideline resulted in a 5% improvement in the reporting quality of surgical case series published in three journals. Further research is needed to identify and successfully navigate existing barriers to greater compliance. Authors, reviewers and editors should adhere to the guidelines to boost reporting quality. Journals should develop their policies and guide for authors to incorporate the guideline and mandate compliance. © 2017 IJS Publishing Group Ltd. Published by Elsevier Ltd. All rights reserved.
Keywords: PROCESS Reporting criteria Case series
1. Introduction
* Corresponding author. E-mail address: [email protected] (M.R. Borrelli).
The surgical case series is a traditional stalwart of the literature. It has been defined as: “an observational study of an uncontrolled group of patients collected or sampled on the basis of a specific disease/
http://dx.doi.org/10.1016/j.ijsu.2017.07.079 1743-9191/© 2017 IJS Publishing Group Ltd. Published by Elsevier Ltd. All rights reserved.
R.A. Agha et al. / International Journal of Surgery 45 (2017) 92e97
exposure/intervention or a specific outcome of interest”. [1] Despite their designation as level four evidence in drive for evidence-based medicine (EBM) over the last few decades, they are still highly prevalent and have a role to play [1]. Recent work by our group found that 25e40% of the research output in 2013 of five surgical specialties published in their top three journals (by impact factor) was a case series [2]. Surgical case series provide important insights and initial evidence on new diseases, the mechanisms of disease and their natural history, rare diseases and rare manifestations and their investigation and management, side effects, drug reactions, the management of complications and reminders of important lessons. They are important in emergency scenarios or with population sub-sets where trials cannot occur e.g. children and those with dementia. Early and initial evidence for new treatments and technologies usually comes in the form of a case series, for example vacuum assisted closure therapy [3] or use of propopanol for haemangiomas [4]. Case series have even been used as a means of putting forth a new classification system, such as Mulliken and Glowacki's work with haemangiomas and vascular malformations [5]. Case series are cheap, economical and easier to conduct than other observational studies such as cohort studies, or randomised controlled trials. Indeed, they may act as a ‘launch pad’ into those types of studies. This is recognised by the IDEAL framework where a form of high quality prospective case series is the basis by which a new technique achieves technical stability prior to multicentre take-up and refinement of its indications. However, case series do suffer with inherent bias and recent systematic reviews have found reporting quality and transparency issues [6]. Our group recently completed a systematic review on the reporting quality of case series in surgery over the period 1990e2014 [7]. From 92 articles that met the inclusion criteria, methodological and reporting issues identified were: failure to use standardised definitions (57%), missing or selective data (66%), poor transparency with incomplete reporting (70%), failure to mention whether alternate study designs were considered (11%) and ‘other’ issues (52%) including the failure to clearly define the patient population under investigation, selection bias insufficient followup time, need for validated outcomes. Readers need complete, clear and transparent information. Authors may benefit from guidelines outlining the minimum criteria necessary to report in a case series to adequate and effectively report their case series. We published the PROCESS guideline in December 2016 to aid the writing of surgical case series [8]. It was developed through a Delphi consensus exercise amongst a 59strong international expert group. At the time of writing, it has been cited 40 times according to Google Scholar, has been downloaded 1286 times, is listed on Equator network website (www. equator-network.org) and has its own dedicated website www. processguideline.com [9]. The impact of the PROCESS guidelines has not been previously assessed. It is important to evaluate the effect PROCESS guidelines have had on reporting case series to assess their utility as a tool that can be used by journal editors and reviewers as guidance, as well as guidance to authors. This study is the first study to assesses the impact of the introduction of the PROCESS guideline on reporting of surgical case series submitted to three international journals. 2. Methods 2.1. Article selection 20 case series published in the International Journal of Surgery Case Reports (IJSCR), the International Journal of Surgery (IJS) or the Annals of Medicine and Surgery (AMS) were assessed in the period
93
September to December 2016. This was prior to the introduction of the PROCESS guideline (the pre-PROCESS period). The articles were randomly identified and scored against the PROCESS criteria, by searching the journal data bases using the key word ‘case series’ in the specified date range. This generated a list of 219 potential articles across the three articles, numbered sequentially for randomisation (N ¼ 69 in IJS, N ¼ 32 in AMS, N ¼ 43 in IJSCR). A computer generated random number between 1 and 219 was used to select the article for inclusion. The article was included provided it was a case series. Articles after the introduction of the PROCESS guideline were only included if they reported adherence to the PROCESS criteria. These three journals were chosen as they were the only ones to have implemented the PROCESS guidelines at that time. 2.2. Article coding Each case series was scored against the 29 items in the PROCESS criteria. An item was given 1 point if it satisfied all criteria in the item, 0.5 if there was partial completion, 0 e for not addressing the item at all and NA (not applicable) if this item did not apply. Each item could score a maximum theoretical ‘PROCESS score’ of 29 (assuming all criteria are applicable). Points would be deducted from this maximum in the denominator if they were deemed not relevant to that specific case series. Table 1 below shows how the total score out of 29 would be reached for the PROCESS criteria: Two independent teams performed scoring. Scores for the two teams were compared and consensus was reached to achieve a final score set. The methodology was revised to involve just two authors performing all scoring. Before scoring the authors discussed each PROCESS point in detail, scored a pilot case series independently and discussed discrepancies to resolve misunderstandings, before independent scoring on the 40 articles was carried out. All scoring was conducted on a pre-formatted excel database. Once scoring was completed ‘PROCESS score’ for each article was calculated and expressed as a percentage. The percentage of each of the 29 PROCESS items answered correctly out of the theoretical maximum of 40 was also calculated. Prior to starting work, data collectors were given a briefing and a standardised data extraction template. Training was provided to the lead data collectors (MB and RF). 2.3. Statistics SPSS was used for statistical analysis (version 23, SPSS Inc, Chicago, IL). The Shapiro Wilks test was used to test for normally distributed data. Although medians and ranges are usually appropriate for descriptive data, means and standard deviations were given for normally distributed data. For normally distributed data the independent-samples t-test was used to compare pre- and post scores (two-tailed). Cohen's Kappa was used to assess agreement between researchers. A score of greater than 0.5 was considered moderate agreement. 3. Results The mean PROCESS score was 79.9% (standard deviation ± 6.9, range 66e90) for the pre-PROCESS period and 84.1% (standard deviation ± 6.2, range 72e95) for the post-PROCESS period, a 5% relative increase. Tests for normality confirmed a normal or Gaussian distribution and this increase was found to be statistically significant (t(19) ¼ 2.164, P ¼ 0.043. This data is summarised in Table 2: The Cohen's Kappa score was 0.907, implying very substantial agreement. Fig. 1 below shows how compliance with the different items changed before and after PROCESS:
94
R.A. Agha et al. / International Journal of Surgery 45 (2017) 92e97
Table 1 The PROCESS guideline checklist and how marks were awarded. Topic
Item
Checklist item description
Title
1
Abstract
2a 2b
The words “case series” and the area of focus should appear in the title (e.g. disease, exposure/intervention or outcome). Introduction - what is the unifying theme of the case series. Methods - describe what was done, how and when was it done and by whom. Results - what was found. Conclusion - what have we learned and what does it mean
2c 2d Introduction
3
Methods
4a
4b
4c
4d
4e
4f
4g
4h
4i
4j
Results
5a
5b
5c
5d
5e
Discussion
6a
How marks were awarded
0.5 points for case series 0.5 points for area of focus 1 point if theme mentioned, else 0 0.5 points for ‘what’ 0.5 points for ‘how’ 1 point for results found, else 0 1 point if mention what learnt, 0 points if not Explain the scientific background and rationale for the case series. What is 0.5 points for referral to literature the unifying theme - common disease, exposure, intervention and outcome, 0.5 points for study need etc. Why is this study needed? Registration and ethics - state the research registry number in accordance 0.5 e for mentioning ethics 0.5 e for registering study with the declaration of Helsinki - “Every research study involving human subjects must be registered in a publicly accessible database” (this can be obtained from; ResearchRegistry.com or ClinicalTrials.gov or ISRCTN). State whether ethical approval was needed and if so, what the relevant judgement reference was? Study design - state the study is a case series and whether prospective or 0.25 - case series retrospective in design, whether single or multi-centre and whether cases 0.25 e if mention prospective/retrospective 0.25 e for mentioning centre are consecutive or non-consecutive. 0.25 e for mentioning whether or not consecutive 1/3 e for mentioning setting where recruit the patients Setting - describe the setting(s)and nature of the institution in which the 1/3 points for mentioning when recruited patient was managed; academic, community or private practice setting? Location(s), and relevant dates, including periods of recruitment, exposure, 1/3 points for mentioning follow-up follow-up, and data collection 1/3 e points for mentioning inclusion/exclusion criteria Participants - describe the relevant characteristics of the participants 1/3 points for mentioning smoking or social health (comorbidities, tumour staging, smoking status, etc). State any eligibility 1/3 points for mentioning comorbidities (inclusion/exclusion) criteria and the sources and methods of selection of participants. Describe length and methods of follow-up. Pre-intervention considerations e.g. Patient optimisation: measures taken 1 point if mention, if applicable, else 0 prior to surgery or other intervention e.g. treating hypothermia/ hypovolaemia/hypotension in burns patients, ICU care for sepsis, dealing with anticoagulation/other medications and so on. Types of intervention(s) deployed and reasoning behind treatment offered 1/3 points for intervention 1/3 points for reasoning (pharmacological, surgical, physiotherapy, psychological, preventive) and 1/3 points for concurrent treatment concurrent treatments (antibiotics, analgesia, anti-emetics, nil by mouth, VTE prophylaxis, etc). Medical devices should have manufacturer and model specifically mentioned. 0.5 points for what was done specifically Peri-intervention considerations - administration of intervention (what, 0.5 points for how this was done where, when and how was it done, including for surgery; anaesthesia, patient position, use of tourniquet and other relevant equipment, preparation used, sutures, devices, surgical stage (1 or 2 stage, etc). Pharmacological therapies should include formulation, dosage, strength, route and duration). 1 point for mentioning who performed the procedure, else 0 Who performed the procedures - operator experience (position on the learning curve for the technique if established, specialisation and prior relevant training). Quality control - what measures were taken to reduce inter or intra1 point for mentioning effort to reduce inter or intraoperator variation. What measures were taken to ensure quality and operator variation consistency in the delivery of the intervention e.g. independent observers, lymph node counts, etc Post-intervention considerations e.g. post-operative instructions and place 0.5 points for mentioning a post-operative care package 0.5 points for mentioning location of follow-up of care. Important follow-up measures - diagnostic and other test results. Future surveillance requirements - e.g. imaging surveillance of endovascular aneurysm repair (EVAR) or clinical exam/ultrasound of regional lymph nodes for skin cancer. Participants - reports numbers involved and their characteristics 1/3 points e for mentioning participant number (comorbidities, tumour staging, smoking status, etc). 1/3 points for mentioning smoking status 1/3 points for mentioning comorbidities Any changes in the interventions during the course of the case series (how 0.5 points for change mentioned if applicable, 0.5 points for why change has it evolved, been tinkered with, what learning occurred, etc) together with rationale and a diagram if appropriate. Degree of novelty for a surgical else NA technique/device should be mentioned and a comment on learning curves should be made for new techniques/devices. 1 point for mentioning follow-up time, else 0 Outcomes and follow-up - Clinician assessed and patient-reported outcomes (when appropriate) should be stated with inclusion of the time periods at which assessed. Relevant photographs/radiological images should be provided e.g. 12 month follow-up. Where relevant - intervention adherence/compliance and tolerability (how 0.5 if lost patients to follow-up was this assessed). Describe loss to follow-up (express as a percentage) and 0.5 points if mention why or, 1 point if no lost to follow up any explanations for it. Complications and adverse or unanticipated events. Described in detail and 0.5 point for mentioning complications 0.5 points for mentioning how managed, ideally categorised in accordance with the Clavien-Dindo Classification. How they were prevented, diagnosed and managed. Blood loss, operative NA if not applicable time, wound complications, re-exploration/revision surgery, 30-day postop and long-term morbidity/mortality may need to be specified. Summarise key results 1 point for summarising results, else 0
R.A. Agha et al. / International Journal of Surgery 45 (2017) 92e97
95
Table 1 (continued ) Topic
Item
Checklist item description
6b
1 point for discussing results with previous literature, else 0 Discussion of the relevant literature, implications for clinical practice guidelines, how have the indications for a new technique/device been refined and how do outcomes compare with established therapies and the prevailing gold standard should one exist and any relevant hypothesis generation. Strengths and limitations of the study 0.5 points for strengths 0.5 points for limitations The rationale for any conclusions? 1 point for rational, else 0 State the key conclusions from the study 1 point for key conclusions, else 0 State what needs to be done next, further research with what study design. 0.5 points for mentioning need for future work, 0.5 points if study design mentioned State any conflicts of interest 1 point if mention, else 0 State any sources of funding 1 point if mention Total ¼ 29
6c
Conclusions
6d 7a 7b
Additional Information
8a 8b
How marks were awarded
Table 2 Summary results for the pre and post PROCESS periods. Period
Mean
S.D.
t
df
P value
Pre-PROCESS (n ¼ 20) Post-PROCESS (n ¼ 20)
79.9 84.1
6.9 6.2
2.053
38
0.047
Abbreviations: SD e standard deviation, S.D. e standard deviation.
From Fig. 1 we can see that the areas that improved the most and these are shown in Table 3 below. Table 4 below shows the top five areas of poorest compliance in the pre and post PROCESS periods. It can be noted that items 4e, 7b and 4h are in the top three reasons for both groups. 4. Discussion Our study showed a 5% relative increase in the compliance with the PROCESS guideline before and after its implementation. In both periods the top three items most poorly complied with were similar (Table 4). The reasons for this consistency is uncertain and was not of specific focus to the research question posed here. In both pre- and post-PROCESS periods point 4h ‘who performed the procedure and their level of experience’ is often missing
from published case series. Authors may feel this is irrelevant, especially for less complex procedures which may be performed by multiple surgeons with different experience levels and credentials. Also, if there was no significant pre-intervention consideration, authors may simply not mention it. The focus of the PROCESS guidelines is to guide authors in how to transparently report both what has and has not been done. The need to report a negative in this context may be underappreciated by authors. The operator and their level of experience, however, can be a strong factor determining surgical outcomes and is therefore important this information is portrayed to readers. Another area poorly reported in case series is point 7b ‘what needs to be done and any future work’. Authors may feel their case series is a definitive guide to practice and hence there isn't a need for further work. However, authors may simply be omitting this information for lack of paying close enough attention to the guidelines. This data is useful going forward as it will provide a focus for future efforts to raise compliance, especially if peerreviewers and editors are aware of these consistent issues. The responsibility for adequate and transparent reporting is not only that of authors but also of the editors and reviewers of journals. The number of reporting guidelines has increased substantially in recent years. The Equator network mentions 362 at the
Fig. 1. Compliance with the 8 PROCESS items before and after the guideline's introduction.
96
R.A. Agha et al. / International Journal of Surgery 45 (2017) 92e97
Table 3 The most items that improved the most following introduction of the PROCESS guideline. Item
Item description
Improvement in compliance
4e
Pre-intervention considerations e.g. Patient optimisation: measures taken prior to surgery or other intervention e.g. treating hypothermia/hypovolaemia/hypotension in burns patients, ICU care for sepsis, dealing with anticoagulation/ other medications and so on. Who performed the procedures - operator experience (position on the learning curve for the technique if established, specialisation and prior relevant training). Study design - state the study is a case series and whether prospective or retrospective in design, whether single or multi-centre and whether cases are consecutive or non-consecutive. The words “case series” and the area of focus should appear in the title (e.g. disease, exposure/intervention or outcome). Registration and ethics - state the research registry number in accordance with the declaration of Helsinki - “Every research study involving human subjects must be registered in a publicly accessible database” (this can be obtained from; ResearchRegistry.com or ClinicalTrials.gov or ISRCTN). State whether ethical approval was needed and if so, what the relevant judgement reference was?
253%
4h 4b 1 4a
135% 51% 47% 42%
Table 4 The top five areas of poorest compliance in the pre and post SCARE periods. Pre-PROCESS (criterion)
Post-PROCESS (criterion)
1
17% - Pre-intervention considerations (item 4e).
2
20% - who performed the procedure (item 4h).
3
35% - State what needs to be done next, further research with what study design (item 7b)
4
55% - Study design - state the study is a case series and whether prospective or retrospective in design, whether single or multi-centre and whether cases are consecutive or non-consecutive. (item 4B)
5
55% - Participants - describe the relevant characteristics of the participants (comorbidities, tumour staging, smoking status, etc). State any eligibility (inclusion/exclusion) criteria and the sources and methods of selection of participants. Describe length and methods of follow-up (item 4D)
40% - State what needs to be done next, further research with what study design (item 7b). 47% - Who performed the procedures - operator experience (position on the learning curve for the technique if established, specialisation and prior relevant training (item 4h). 60% - Pre-intervention considerations e.g. Patient optimisation: measures taken prior to surgery or other intervention e.g. treating hypothermia/ hypovolaemia/hypotension in burns patients, ICU care for sepsis, dealing with anticoagulation/other medications and so on (item 4e) 63% - Setting - describe the setting(s)and nature of the institution in which the patient was managed; academic, community or private practice setting? Location(s), and relevant dates, including periods of recruitment, exposure, follow-up, and data collection (item 4c) 67% - Participants - reports numbers involved and their characteristics (comorbidities, tumour staging, smoking status, etc) (item 5a)
time of writing (July 2017) [10]. Surgical journals have been slow to take up reporting guidelines. Previous research has shown how the majority (62%) of the 193 surgical journals listed in the surgery category in the Journal Citation Report 2014, made no mention of reporting guidelines within their guide for authors at all [11]. Evidence on the effectiveness of reporting guidelines is mixed. A Cochrane systematic review focusing on randomised controlled trials and CONSORT was conducted by Turner et al. [12] It found that 25 of 27 outcomes assessing completeness of reporting favoured CONSORT-endorsing journals over non-endorsers, of which five were statistically significant. Simply endorsing a guideline doesn't mean that 100% compliance will be achieved and our data support this Cochrane review. In IJS, AMS and IJSCR all authors are required to submit a completed checklist for case series as well as other common manuscript types and this is available to peerreviewers and the editor. Authors are unable to submit a manuscript until they have met this requirement. The magnitude of effect (i.e. a 5% increase) is similar to what we have seen for other guidelines where submission of a completed checklist was made mandatory as part of a wider policy initiative [13]. Further research is needed to understand how we can increase compliance yet further from the 84% achieved here. Solutions may be more forthcoming once more journals adopt the guideline and assess their own data on compliance. There are several limitations of our study. Firstly, these analyses were conducted soon after the guidelines were published in December 2016 and hence awareness will be more limited and authors, reviewers and editors will still be getting used to them. A
follow-up study further down the line is planned which will be a useful additional comparison. The analysis is comparing only one pre-PROCESS and one post-PROCESS period, so chance and Hawthorne effects can play a role. The journals chosen are only three of the many journals in which case series are published, so generalisability and external validity is limited somewhat. Reviewers were not blinded to the articles (authors would mention their work was compliant with the PROCESS criteria in the text so this was not feasible) and articles were not reviewed in randomised order (i.e. pre-PROCESS first and then post-PROCESS) and there is a degree of subjectivity in marking.
5. Conclusion Implementation of the PROCESS guideline resulted in a 5% improvement in the reporting quality of surgical case series published in three journals. Authors, reviewers and editors should adhere to the guidelines to boost reporting quality and journals should develop their policies, submission processes and guide for authors to incorporate the guideline. Other journals publishing surgical case reports should consider utilising the PROCESS guideline and how they can achieve compliance with it during the process of submission and peer-review.
Ethical approval Not required for this before after study.
R.A. Agha et al. / International Journal of Surgery 45 (2017) 92e97
Sources of funding None. Author contribution RAA developed the concept, analysed and interpreted the data, wrote the first draft and revised it in line with feedback from other authors. TKO, MMC, RF, MRB, RT and JG were involved in data collection. MRB conducted the statistical analyses. ND and DO was involved in data interpretation. All authors were involved in critically reviewing the manuscript draft and approving the final version for submission. Conflicts of interest None. Research registration unique identifying number (UIN) This is not a human study. Guarantor RAA. References [1] Oxford Centre for Evidence-based Medicine e Levels of Evidence, March 2009 [online]. Available at, http://www.cebm.net/oxford-centre-evidence-basedmedicine-levels-evidence-march-2009/. Accessed 16 July 2017.
97
[2] R.A. Agha, M. Devesa, K. Whitehurst, A.J. Fowler, R. Coe, G. Wellstead, D.P. Orgill, P. McCulloch, Levels of evidence in plastic surgery - bibliometric trends and comparison with five other surgical specialties, Eur. J. Plastic Surg. 39 (5) (2016) 365e370. [3] L.C. Argenta 1, M.J. Morykwas, Vacuum-assisted closure: a new method for wound control and treatment: clinical experience, Ann. Plast. Surg. 38 (6) (1997 Jun) 563e576. aute -Labre ze, E. Dumas de la Roque, T. Hubiche, F. Boralevi, J.B. Thambo, [4] C. Le A. Taïeb, Propranolol for severe hemangiomas of infancy, N. Engl. J. Med. 358 (24) (2008 Jun 12) 2649e2651, http://dx.doi.org/10.1056/NEJMc0708819. [5] J.B. Mulliken, J. Glowacki, Hemangiomas and vascular malformations in infants and children: a classification based on endothelial characteristics, Plast. Reconstr. Surg. 69 (3) (1982 Mar) 412e422. [6] R.A. Agha, A.J. Fowler, C. Herlin, T. Goodacre, D.P. Orgill, Use of autologous fat grafting for breast reconstruction: a systematic review and meta-analysis, J. Plastic Reconstr. Surg. 68 (2015) 143e161. [7] R. Agha, A. Fowler, S. Lee, B. Gundogan, B. Whitehurst, H. Sagoo, K.J.L. Jeong, D. Altman, D. Orgill, A systematic review of the methodological and reporting quality of care series in surgery, Br. J. Surg. 103 (10) (2016) 1253e1258. [8] R.A. Agha, A.J. Fowler, S. Rajmohan, I. Barai, D.P. Orgill, PROCESS Group, Preferred reporting of case series in surgery; the PROCESS guidelines, Int. J. Surg. 36 (Pt A) (2016) 319e323. [9] Google Scholar: The SCARE statement: consensus-based surgical case report guidelines [online]. Available at: https://scholar.google.com/citations?view_ op¼view_citation&hl¼en&user¼Ig4lUDEAAAAJ&citation_for_ view¼Ig4lUDEAAAAJ:WA5NYHcadZ8C (Accessed 16 July 2017). [10] Equator network [online]. Available at: http://www.equator-network.org/. [11] R.A. Agha, I. Barai, S. Rajmohan, S. Lee, M.O. Anwar, A.J. Fowler, D.P. Orgill, D.G. Altman, Support for reporting guidelines in surgical journals needs improvement: a systematic review, Int. J. Surg. (2017) article in press. [12] L. Turner, L. Shamseer, D.G. Altman, L. Weeks, J. Peters, T. Kober, et al., Consolidated standards of reporting trials (CONSORT) and the completeness of reporting of randomised controlled trials (RCTs) published in medical journals, Cochrane database Syst. Rev. 11 (2012 Jan). MR000030. [13] R. Agha, A. Fowler, C. Limb, K. Whitehurst, R. Coe, H. Sagoo, D. Jafree, C. Chandrakumar, B. Gundogan, Impact of the mandatory implementation of reporting guidelines on reporting quality in a surgical journal: a before and after study, Int. J. Surg. 30 (2016) 169e172.