What types of intervention have been implemented in order to influence self-administration among hospital inpatients?
What effects have these interventions had on outcomes related to patients (e.g. adherence to medication regimens, medication errors, medicines knowledge, satisfaction with SAM programmes and success in self-medication), staff (e.g. satisfaction with SAM programmes, staff time) and on the institution (e.g. costs vs. savings)?
We wish to conduct a systematic search for published texts using the following databases: EMBASE (1980-2012), MEDLINE (1950-2012), CINAHL (1981-2012), PsycINFO (1967-2012), NELM, HMIC, CPCI (Web of Knowledge) and Zetoc.
We will also carry out a targeted internet search to identify grey literature to overcome publication bias. We will undertake reference harvesting and citation searching of key documents identified. There are to be no country or language exclusions.
Types of study to be included
We envisage drawing on a broad range of study types including both quantitative and qualitative evaluative studies. Depending on the volume studies returned from searches we will focus first on the following types of studies: Controlled trials, cohort studies (controlled before and after studies), interrupted time series and case-series (uncontrolled before and after studies). We plan to select studies for inclusion only if they include a description of their intervention and formal evaluation of this, either qualitative or quantitative. Purely anecdotal or descriptive papers will be excluded. In order to avoid publication bias, any relevant reports, conference abstracts, or theses will be included, regardless of publication status, providing that adequate information concerning study design and outcomes is reported.
Condition or domain being studied
Hospitals in which a self-administration of medicines programme was implemented and the effects measured. Community hospitals will be included, but not home self-administration or residential homes. We wish to investigate environments that are primarily medical, rather than a home with nursing staff present. We will not look at the impact that self-administration programmes have on medical staff, but will take into account the perceptions of the implementation and effectiveness of self-management programmes where these are robustly recorded.
Hospital inpatients. We will exclude studies that exclusively consider outpatients. Hospital inpatients that are discharged forming part of follow-up studies may be included.
Any intervention or combination of interventions which aimed to encourage self-administration among hospital inpatients.
None, or usual care where defined.
Any inpatient setting where an intervention has been implemented in order to encourage patient self administration of prescribed medicines. To avoid language bias, publications written in any language will be considered for review, providing a translation can be obtained.
Adherence to medication regimens within hospital, and following discharge if recorded.
Errors in drug administration.
Patient experiences of participation in self-administration programmes (e.g. patient satisfaction, standardised PROMS (patient reported outcome measures) and staff experiences of involvement in the scheme (e.g. perceived efficacy, unintended consequences).
Data extraction, (selection and coding)
Two reviewers will independently complete each phase of the study selection and data extraction process. Disagreements will be resolved via discussion. If a joint decision cannot be reached, a third reviewer will act as arbitrator.
All citations identified through the searches and cross-referencing will be entered into a reference manager software library. Selection of studies for inclusion in the review will be completed in two stages.
(1) After the removal of duplicates, the title and abstract of each citation will be scanned for relevance against the inclusion/exclusion criteria. Irrelevant titles and abstracts will be excluded. A pre-piloted form will be used to document this process.
(2) The final stage involves assessing the full text of each citation. Again, a pre-piloted form will be used to guide and document the assessment of the full text against the inclusion and exclusion criteria.
Those citations that are deemed relevant for inclusion at this stage will be included in the review and subjected to data extraction.
Systematic and consistent data extraction will be ensured by the use of a pre-piloted data extraction form which will include:
• Study authors, title, date of publication.
• Study aims as stated.
• Patient population description and setting; recruitment strategy and how many patients completed (and reasons for attrition if applicable).
• Nature of Intervention (Rationale; description of intervention; relevant contextual information).
• Information on baseline and comparators.
• Type of study, description of study design, planned and actual data analysis.
• Primary outcomes and secondary outcomes (measures used, baseline and subsequent values, length of follow-up, notes on interpretation on scale data where used).
• barriers/facilitators identified.
• Reported outcomes.
• Authors conclusions.
• Limitations mentioned by authors.
• Contact details for primary author and points for clarification.
• Funding and declared conflicts of interest.
Study authors may be contacted and asked to provide further information, especially where this will facilitate inclusion of study data in a meta-analysis. In such cases, emails requesting the information will be sent to the corresponding author using the email address listed in the publication.
Risk of bias (quality) assessment
Quality assurance of studies will be performed separately by two reviewers using the most appropriate quality assessment tool for the type of study. For example for RCTS we will use the Cochrane Risk of Bias tool.
Strategy for data synthesis
A narrative synthesis will be adopted in order to combine the differing study types likely to be found (based on an initial search of the literature and previous systematic reviews on the topic of self-administration).
We will group papers by type of study/methodology deployed for analysis purposes – i.e. separate out RCTs that look at efficacy from more qualitative/mixed methods that look at implementation (barriers/facilitators/drivers/contingent factors) or attempt to identify good practice. The types of papers subject to review will influence the approach for synthesising papers – depending on the quality of RCTs it may be possible to do a meta analysis using commonly reported outcome measures. Depending on the richness of the data reported in more qualitative studies, we will adopt either a realist (context-mechanism-outcome) or thematic analysis approach.
A matrix approach will be adopted in which studies will be grouped by study design and outcomes reported. Where possible, meta-analyses on most commonly reported outcomes will be completed (e.g. medication regimen adherence rates) using random effects model.
Analysis of subgroups or subsets
None currently planned.
Findings will be shared amongst practitioners and if appropriate interim findings will be sent to practitioners for the development of practical recommendations. We plan to produce an article to be published in a peer-reviewed journal. We also aim to produce a CLARHC bite (leaflet to inform the public of research completed within the CLARHC research project).
Contact details for further information
School of Clinical and Experimental Medicine
College of Medical and Dental Sciences (CMDS)
Medical School Building
University of Birmingham
Edgbaston B15 2TT
Organisational affiliation of the review
University of Birmingham
Dr Jamie Coleman, University of Birmingham Mrs Suzanna Richardson, University Hospitals Birmingham NHS Foundation Trust Miss Hannah Brooks, University Hospitals Birmingham NHS Foundation Trust Mr George Bramley, University of Birmingham
Anticipated or actual start date
07 January 2013
Anticipated completion date
31 May 2013
Conflicts of interest
Subject index terms status
Subject indexing assigned by CRD
Subject index terms
Hospitalization; Humans; Patient Compliance; Self Administration; Self Care
Stage of review
Date of registration in PROSPERO
07 January 2013
Date of publication of this revision
04 September 2013
Stage of review at time of this submission
Piloting of the study selection process
Formal screening of search results against eligibility criteria
Risk of bias (quality) assessment
PROSPERO This information has been provided by the named contact for this review. CRD has accepted this information in good faith and registered the review in PROSPERO. CRD bears no responsibility or liability for the content of this registration record, any associated files or external websites.