Objectives

• To systematically identify studies that evaluate interventions aimed at improving quality of surgical and anaesthesia care at hospital level in SSACs.

• To capture, categorise and assess the study characteristics (setting and intervention).

• To capture and categorise all outcomes reported under the following categories: clinical, process, implementation or other.

• To assess methodological rigour of studies with mortality as a primary outcome (which is defined as mortality within 30 days defined for this review).

• To record whether process and implementation outcomes were assessed.

Search strategy

The search string strategy was developed by the main author and checked by a research librarian for errors and omissions. During the development process, the search was tested and retested to ensure optimal sensitivity and specificity and search terms refined accordingly. The search strategy was developed in MEDLINE (online supplementary appendix 1). The search terms used intended to cover a wide range of studies that evaluated interventions aimed at improving the quality of surgical and anaesthesia care at hospital level in SSACs. The three distinct search strings that were combined to identify these studies are (1) surgical and anaesthesia care, (2) quality improvement hospital-based interventions and (3) SSACs.

Surgical and anaesthesia care will include all branches of surgery (cardiothoracic, ear, nose and throat (ENT), general, neonatal, obstetric, ophthalmic, orthopaedic, paediatric and urology), including operative and non-operative aspects of care, on the presumption that any branch of surgery may have relevant transferable interventions within a low resource setting. For the purpose of this review only studies that explicitly refer to a surgical procedure or a surgical patient will be included. Trauma (both explicitly focusing and included as part of the intervention) will be excluded, as it will be covered in a separate systematic review.

For the purposes of this review health system strengthening intervention (HSSI) is defined as any actions or strategies taken to improve the quality of healthcare delivery or patient outcomes. This will include studies that either strengthen the whole surgical system within a hospital or contribute to one aspect of surgical and anaesthesia care. The WHO Health System Framework will be used to categorise HSSIs.11

Hospital-based interventions are defined as any intervention that is delivered within a tertiary/teaching/national or a secondary/district/regional care hospital environment, aimed at improving quality of care, that directly or indirectly involves care delivery to patients or by staff. Both hospital-wide (hospital documentation, early warning scores, intensive care outreach) and individual departmental (emergency department, surgical wards, theatre, high dependency unit (HDU), intensive treatment unit (ITU), outpatient, radiology) interventions will be included.

The term SSACs will include all countries listed as SSA counties by the World Bank for the 2019 fiscal year.12

To combine terms within each search string, the Boolean operator ‘OR’ was used. To incorporate all variations of the same term, truncation and wildcards were employed. Medical Subject Headings terms and free text were also used. To focus the search on these required terms, phrase and adjacency searching, using proximity operator filters and limits, was exploited. Any additional key terms identified during the literature search were added to the appropriate string to ensure it was maximally inclusive. The three strings were then combined using the Boolean operator ‘AND’.

This search strategy will be adapted to EMBASE, Global Health, CINAHL and Web of Science databases. A reference list of all included articles will be hand-searched for further relevant studies.

To identify unpublished studies, the Scopus literature databases will be searched. In addition, first 50 hits from Google search, documentation and reports of relevant national and international health organisations such as Comic Relief, DFID, Grand Challenges Canada, the Bill and Melinda Gates Foundation, HIVOS (Humanistisch Instituut voor Ontwikkelingssamenwerking), World Vision, the Robert Wood Johnson foundation, United Nations Children’s Fund (UNICEF) and GOAL will be searched.

The choice of information sources for this search were made based on the recommendation from the current literature, to ensure the optimal combination of databases and other search resources available needed to guaranty adequate and efficient coverage.13

The search process will be recorded using a Database literature search log (online supplementary appendix 2).

Eligibility criteria

All eligible articles published in English will be included, from 2008 to 2019. The inclusion and exclusion criteria are illustrated in table 1 below.

Data extraction

All potential articles will be initially collated in Endnote X8 for de-duplication of results. All remaining studies will be loaded into Rayyan online opensource web application14 and initially screened by title and abstract by a team of two reviewers independently, according to the inclusion and exclusion criteria (table 1) and by using the Systematic Review Abstract and Full-Text Screening Criteria form (online supplementary appendix 3). All reviewers were appropriately trained in systematic review methods. If either of the reviewers include an abstract, it will be promoted to consideration for full-text review. The full-text review will be done by the first author (NB) and another reviewer independently, using inclusion and exclusion criteria. Data elements will be extracted into a predefined data extraction form (online supplementary appendix 4), which will be piloted before finalising the elements to be included. This includes information on authors, publication dates, country of intervention, healthcare setting, type of intervention, methodology or design, patient/intervention size, clinical/process/implementation/other outcomes and the impact of the intervention. The data extracted by two reviews will be compared and if any discrepancies arise, consensus will be sought among the study team (figure 1).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Data search and extraction flow diagram.

Outcomes

Three type of outcomes, clinical, process and implementation, will be collected. The primary outcome for this review will be mortality (in-hospital perioperative mortality and post intervention mortality within 30 days). All other outcomes will be secondary: clinical outcome such as major and minor complications as defined by the Clavien-Dindo Classification (table 2) and any other clinical outcomes identified during the course of reviewing the manuscripts. Process outcomes will include all outcomes relevant to interventional studies included in this review (waiting times, length of hospital stay, blood availability, and so on). Implementation outcomes will be determined using a ‘working taxonomy’ of eight conceptually distinct implementation outcomes produced by Proctor et al15 (table 3). If any other outcomes are reported that cannot fit into the three categories above, they will be recorded under the heading ‘other outcomes’, described, and categorised, if possible, after discussion among authors.

Risk of bias

Risk of bias of individual studies and publication bias will be addressed by designing a thorough search strategy to find relevant studies. The search strategy will be assessed by checking the results against known articles which are relevant. To address publication bias the grey literature will be searched, as described above, to include all relevant studies which are not reported in peer-reviewed journals. The following type of publications will be included: dissertations/theses, books/book chapters, conference abstracts and research in progress. Also, all potential full-text articles will be screened by two reviewers. For studies that reported their main outcome as mortality the reviewers will independently assess the risk of bias as part of the Grading of Recommendations Assessment, Development, and Evaluation (GRADE) system16 of rating quality of evidence and grading strength of recommendations in systematic reviews.

Disagreements between the reviews over the risk of bias level (serious or very serious) will be resolved by discussion with involvement of a third reviewer where necessary. Due to expected high heterogenicity of the data extracted, a meta-analysis will not be feasible.

Data synthesis

For all outcomes all items that will be collected using predefined collection form (online supplementary appendix 4) will be summarised and the methodological approach used in the studies will be identified. Descriptive statistics and narrative synthesis of data will be undertaken to appreciate the diversity of research approaches that are used for assessment of quality of surgical and anaesthesia care within hospital. Those studies that were collected under the heading ‘other’, we will look for common themes and then compare these themes to our three defined categories of outcomes.

In addition, the studies reporting mortality as the main outcome of the interventions will be assessed using GRADE criteria for rating quality of evidence and grading strength of recommendations in systematic reviews.16 The criteria are as follows: study design, risk of bias, imprecision, inconsistency, indirectness, and magnitude of effect. They will be summarised in summary of findings table (online supplementary appendix 5). Based on this information quality of evidence will be rated and detailed information about the reason for the quality of evidence rating will be recorded. Outcomes will be categorised as strong or weak according to the quality of the supporting evidence.

Patient and Public Involvement

No patient involved.

Ethics and dissemination

This study is a part of a wider National Institute of Health Research (NIHR) Global Health Research Unit on Health System Strengthening in SSA (ASSET), King's College London (GHRU 16/136/54) project. The results will be communicated and disseminated during meetings with key stakeholders at national, district and community levels, that will be on-going throughout the project. The results will inform the development and implementation of a quality improvement intervention to improve quality of surgical and anaesthesia care in SSACs in general and in Sierra Leone in particular. The results of the review will be published in a peer reviewed journal and presented at national and international conferences. As this is a review of secondary data no formal ethical approval is required.

To our knowledge this is the first systematic literature review to synthesise the information on HSSIs to improve the quality of hospital-based surgical and anaesthesia care in SSACs. The findings from this systematic review will provide evidence to inform the development, design and implementation of the HSSIs to improve the quality of health care for surgical patients at hospital level in SSACs. It will also provide valuable information for policy makers, healthcare managers and clinicians.