The initial appointment non-attendance within mental health services. This problem has important economic and clinical implications as fewer people successfully access help and at the same time resources within mental health services are wasted.
Adult referrals by any third party to a mental health service.
Adults referred to group therapy.
Adults attending an 'initial' outpatient appointment after discharge from inpatient care.
Any intervention with the explicit intention to increase initial appointment attendance or reduce initial appointment non-attendance.
Interventions that involve an appointment at the referred-to service.
Routine procedure or any active comparator.
Attendance rates of the initial appointments.
The outcome will be calculated on intention-to-treat basis, where the risk in each group is calculated as the number of people who attend their initial appointment over the total number randomized to that group.
Studies with all time periods between referral and the appointment will be included but this will be recorded for each study. If the study does not report this, attempts will be made to contact the authors to obtain this information.
If meta-analyses are undertaken, the heterogeneity will be investigated through subgroup analyses. The studies will be divided in terms of short- (1-4 weeks), medium- (4-8 weeks) and long – term (8 weeks or longer) outcomes.
Number of people who cancelled their appointments in the intervention and control group.
The risk of cancelling the appointment in each group is calculated as the number of people who cancelled their appointment over the total number of people randomized to that group.
The number of people who rescheduled their appointments in each group.
If the attendance of people who rescheduled their appointments is not included in the main outcome, this will be recorded to test if there was a difference in rates of rescheduling between the groups.
Data extraction, (selection and coding)
Selection of studies:
Titles and abstracts that are identified through electronic searching will be screened by the review team to identify if they are potentially relevant for this review. Two reviewers will independently screen the titles and abstracts and mark them as “not relevant” or “potential”. In case of any disagreements, the study will be marked as “potential”.
Full reports will thereafter be obtained for all of the studies marked as potentially relevant. After this, two reviewers will independently apply the inclusion and exclusion criteria to the obtained studies. The reviewers will discuss any disagreements and, if no agreement can be reached, a third researcher will decide if the study is to be included. If no decision can be reached through discussion, authors will be contacted for further information. If meta-analyses are undertaken, the effect of studies included through third-author decision will be investigated in sensitivity analyses.
Data extraction and management:
A data extraction form will be piloted on three studies that fit the inclusion criteria and modified if necessary by one of the reviewers. The data will thereafter be extracted independently by two of the reviewers and entered on to a spreadsheet. In the case of any disagreements, an expert statistician will be consulted and, if necessary, and the authors will be contacted for further information.
Risk of bias (quality) assessment
Assessment of the risk of bias:
The Cochrane Collaboration risk of bias tool (Higgins and Altman, 2008) includes the following domains:
1. Sequence generation.
2. Allocation concealment.
3. Blinding of participants, personnel and outcome assessors.
4. Incomplete outcome data.
5. Selective outcome reporting.
6. Other sources of bias.
For each of the included studies, two reviewers will independently rate each domain as “Yes”, “No” or “Unclear”. Any disagreements will be discussed in the review team and the authors will be contacted for further information if necessary. If consensus cannot be reached, the description “unclear” will be used.
In case meta-analyses are undertaken, the potential effect of bias will be investigated in sensitivity analyses.
Strategy for data synthesis
Assessment of heterogeneity
The review will attempt to consider all the studies in each intervention category together. As the range of different interventions is not yet known, the categories will be constructed post hoc. As the delivery of interventions is likely to vary between studies, clinical heterogeneity is expected. This will be discussed in the review team and thereafter explored statistically.
A chi-squared statistic will be obtained to assess the presence of variation between studies that is not due to chance alone. As this test has low power if only a small number of studies are included in the analyses, the alpha-level will be set at 0.1.
An I-squared statistic will also be obtained to investigate the degree of variability due to heterogeneity. If the chi-squared is significant and/or the I-squared value lies between 50-90%, substantial statistical heterogeneity is likely to be present and it will be investigated further through subgroup analyses. The likely sources of clinical heterogeneity are: the referral route of the participants, comparison group used and timing of the outcome assessment.
If the I-squared statistic is <50% and the synthesis of the studies is deemed to be clinically meaningful, both the fixed effects and random effects models will be applied to the data to investigate the effect of the residual heterogeneity. If meta-analyses are deemed inappropriate (e.g., because of a small number of studies identified or high degree of heterogeneity) a narrative synthesis approach will be chosen.
Narrative synthesis plan:
In case a narrative synthesis is deemed most appropriate, the characteristics of the included studies will be summarised in a table format. Any common themes or trends identified across studies will also be discussed in text.
Analysis of subgroups or subsets
1. The studies will be divided into groups by referral routes. The groupings will be: psychiatrist, GP, self-referral and other.
2. The studies will be divided by the comparison intervention used. The groupings will be: only appointment given, appointment and appointment letter, appointment and a reminder and opt-in system.
3. The studies will be divided into groups by the average amount of time between the referral and the initial appointment. The groupings will be: short- (1-4 weeks), medium- (4-8 weeks) and long – term (8 weeks or longer).
The completed review will be submitted to a leading journal in the field.
A comprehensive report will also be made available free of charge on the Institute of Psychiatry online database.
Contact details for further information
Health Service and Population Department
De Crespigny Park, London
United Kingdom, SE5 8AF
Organisational affiliation of the review
Institute of Psychiatry, King's College London
Mr Oliver Schauman, Institute of Psychiatry Dr Sarah Clement, Institute of Psychiatry
Anticipated or actual start date
15 June 2011
Anticipated completion date
31 January 2012
Conflicts of interest
Subject index terms status
Subject indexing assigned by CRD
Subject index terms
Appointments and Schedules; Humans; Mental Health Services; Patient Compliance
Stage of review
Date of registration in PROSPERO
10 June 2011
Date of publication of this revision
19 December 2011
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.