Tipos de estudios

Se incluyeron todos los ECA que evaluaron la TPC en comparación con la TCCCT o una condición control. No se utilizó el contexto, el tamaño de la muestra ni el estado de publicación para determinar la inclusión de los estudios.

Tipos de participantes

Tipos de intervenciones

Tipos de medida de resultado

Se incluyeron los estudios que cumplieron con los criterios de inclusión mencionados, independientemente de si informaron de los resultados siguientes.

Cochrane Common Mental Disorders Controlled Trials Register (CCMD‐CTR)

The Cochrane Common Mental Disorders Group retains a specialized register of RCTs ‐ the CCMD‐CTR. This Register contains over 40,000 reference records (reports of RCTs) for anxiety and depressive disorders, bipolar disorder, eating disorders, self‐harm, and other mental disorders within the scope of this Group. The CCMD‐CTR is a partially studies‐based register with more than 50% of reference records tagged to about 12,500 individually PICO‐coded study records. Reports of trials for inclusion in the Register are collated after (weekly) generic searches of MEDLINE (1950‐), Embase (1974‐), and PsycINFO (1967‐); quarterly searches of the Cochrane Central Register of Controlled Trials (CENTRAL); and review‐specific searches of additional databases. Reports of trials are also sourced from international trial registries, drug companies, key journals (upon handsearching), conference proceedings, and other (non‐Cochrane) systematic reviews and meta‐analyses. Details of CCMD's core search strategies (used to identify RCTs) can be found on the Group website; an example of the core MEDLINE search is displayed in Appendix 1. The register is current to June 2016 only.

Búsquedas electrónicas

CCMD's Information Specialist searched the CCMD‐CTR to 1 June 2016, as follows:

• Cross‐search of the studies and reference registers using the following terms to identify relevant reports on RCTs: (present centred or present centered or present focused or present focussed).

As the CCMDCTR was only current to June 2016, CCMD's Information Specialist ran additional searches on the following databases in February 2018 and February 2019 (Appendix 2):

• Ovid MEDLINE (1946 to 15 February 2019);

• Ovid Embase (1974 to 2019 Week 07);

• Ovid PsycINFO (1806 to February Week 1 2019);

• Cochrane Central Register of Controlled Trials (CENTRAL) (all years to Issue 2, February 2019);

• WHO ICTRP (all years to 15 February 2019);

• Clinicaltrials.gov (all years to 15 February 2019);

• ProQuest PTSDpubs (all years to 15 February 2019);

• ProQuest Dissertations & Theses Global (all years to 15 February 2019).

Two review authors screened all references to check for eligibility. When appropriate, we tagged reports of the same trial together to ensure that no trial was counted twice.

We applied no restrictions based on date, language, or publication status to the searches.

We also searched international trial registries via the trials portal of the World Health Organization (ICTRP) and ClinicalTrials.gov, to identify unpublished and ongoing studies.

We searched reference lists of included studies for additional relevant studies and screened other systematic reviews of psychological interventions for PTSD to identify additional studies not retrieved by our search.

Búsqueda de otros recursos

Selección de los estudios

Two review authors (BB, EB) independently screened titles and abstracts for potential inclusion of all studies identified as a result of the search and coded them as 'retrieve' (eligible or potentially eligible/unclear) or 'do not retrieve.' We retrieved full‐text study reports/publications, and two review authors (BB, EB) independently screened full texts to identify studies for inclusion, and to identify and record reasons for exclusion of ineligible studies. We resolved disagreements through discussion, or, if required, we consulted a third review author (DE). We identified and excluded duplicate records and collated multiple reports that related to the same study, so that each study rather than each report was the unit of interest in the review. We recorded the selection process in sufficient detail to complete a PRISMA flow diagram and a 'Characteristics of excluded studies' table.

Extracción y manejo de los datos

We used a data collection form that was piloted on at least one study in the review to extract study characteristics and outcome data. Two review authors (BB, EB) extracted the following study characteristics and outcome data from included studies.

• Methods: study design, study duration, study setting, recruitment, number of study centers and locations, withdrawals, and dates of study.

• Participants: N, mean age, age range, gender, severity of condition, trauma type, duration of time since trauma, comorbid conditions, diagnostic criteria, inclusion criteria, and exclusion criteria.

• Interventions: interventions and comparisons.

• Outcomes: primary and secondary outcomes specified and collected, method of collection, and time points reported.

• Notes: funding for trial and notable conflicts of interest of trial authors.

We noted in the Characteristics of included studies table if outcome data were not reported in a usable way. We resolved disagreements by consensus or by consultation with a third review author (DE). One review author (EB) transferred data into the Review Manager (RevMan 2014) file. A second review author (BB) double‐checked that data were entered correctly by comparing data presented in the systematic review with data provided in the study reports.

Evaluación del riesgo de sesgo de los estudios incluidos

Two review authors (BB, EB) independently assessed risk of bias for each study using the criteria outlined in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011). We resolved disagreements by discussion or by consultation with another review author (DS). We assessed risk of bias according to the following domains. If information was not reported in the trial publications, then authors were contacted and this information was requested.

• Random sequence generation: describes the method used to generate the allocation sequence in sufficient detail to allow an assessment of whether it should produce comparable groups. We considered this domain to be at 'low' risk of bias if investigators described a process by which each participant had an equal chance of being randomized to each group; at 'high' risk of bias if investigators describe a non‐random component in the sequence generation process; and at 'unclear' risk of bias if information was insufficient for a judgment of high or low risk of bias.

• Allocation concealment: describes the method used to conceal the allocation sequence in sufficient detail to determine whether intervention allocations could have been foreseen in advance of, or during, enrolment. We considered this domain to be at 'low' risk of bias if there was no chance of investigators foreseeing participant assignment; at 'high' risk of bias if investigators could possibly foresee assignments, such as allocation based on alternation or rotation, or an open random allocation schedule (e.g. a list of random numbers); and at 'unclear' risk of bias if information was insufficient for a judgment of high or low risk of bias.

• Blinding of participants and personnel: describes all measures used, if any, to blind study participants and personnel from knowledge of which intervention a participant received. When administering psychological interventions, it is not feasible to blind participants or personnel administering the intervention.

• Blinding of outcome assessment: describes all measures used, if any, to blind outcome assessors from knowledge of which intervention a participant received. We considered lack of blinding separately for patient‐reported and clinician‐rated outcomes. We considered this domain to be at 'low' risk of bias for an outcome if outcome assessors were blinded; at 'high' risk of bias for an outcome if outcome assessors were not blinded; and at 'unclear' risk of bias for an outcome when information was insufficient for a judgment of high or low risk of bias. For self‐report measures, we rated all outcomes as 'high'.

• Incomplete outcome data: describes the completeness of outcome data for each main outcome, including attrition and exclusions from analysis. We considered this domain to be at 'low' risk of bias for an outcome if no data were missing, or if data were imputed appropriately; at 'high' risk of bias for an outcome if missing data were likely related to true outcomes, if analyses considered only the data of treatment completers, or if missing data were imputed inappropriately; and at 'unclear' risk of bias for an outcome when information was insufficient for a judgment of high or low risk of bias. In sum, we considered the effect of incomplete outcome data separately for each outcome and took into account whether reasons for missing data were acceptable, whether trial authors conducted an intention‐to‐treat (ITT) analysis, and the potential impact of missing data on the particular outcome.

• Selective outcome reporting: describes how the possibility of selective outcome reporting was examined by the review authors, and what they found. We considered this domain to be at 'low' risk of bias if the study protocol was available and all prespecified outcomes were reported, or if a study protocol was not available but it is clear that published reports included all expected outcomes; at 'high' risk of bias if not all of the study's prespecified outcomes were reported, if they were reported incompletely, or if primary outcomes were not prespecified or outcomes of interest were reported incompletely; and at 'unclear' risk of bias if information was insufficient for a judgment of high or low risk of bias.

• Other bias: describes any important concerns about bias not addressed by the other domains in the tool but arising during our assessment process. We judged other risks of bias as 'low' or 'high' based on their threats to validity. We considered this domain to be at 'unclear' risk of bias if information was insufficient for a judgment of high or low risk of bias.

Medidas del efecto del tratamiento

Cuestiones relativas a la unidad de análisis

Manejo de los datos faltantes

We contacted investigators or study sponsors to verify key study characteristics and to obtain missing numerical outcome data when possible (e.g. when a study was identified as abstract only). We documented all correspondence with trialists and reported which trialists responded.

We did not use data from an outcome measure when more than 50% of data were missing. For continuous outcomes, we calculated missing standard deviations from other available data, such as confidence intervals, standard errors, and P, T, or F values. As we used only summary measures for the analysis, we assumed that missing data for each study were randomly distributed and thus did not influence the quality of estimates.

Evaluación de la heterogeneidad

Studies brought together in a systematic review will inevitably present various types of heterogeneity, conventionally classified as clinical, methodological, and statistical heterogeneity in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011). In assessing clinical heterogeneity, we closely examined differences between factors associated with intervention or participant characteristics across studies included in the meta‐analysis. When inspecting methodological heterogeneity, we identified differences between methodological factors across studies that could result in substantial diversity in outcome measurements. We paid particular attention to whether outcome variables were defined in the same fashion; whether they were measured by the same quantity and scale; and whether, if differences did exist, methodological diversity affected the quality of the summary effect size estimate. When clinical and methodological heterogeneity does occur in meta‐analysis, statistical heterogeneity becomes inevitable. We measured the degree of inconsistency in findings across studies by using the I² statistic (Higgins 2003). Specifically, the I² statistic indicates the percentage of observed variation that is attributed to true differences across studies; accordingly, we interpreted the I² score by adhering to the following criteria, as proposed in the Cochrane Handbook for Systematic Review of Interventions (Higgins 2011).

• 0% to 40%: might not be important.

• 30% to 60%: may represent moderate heterogeneity.

• 50% to 90%: may represent substantial heterogeneity.

• 75% to 100%: considerable heterogeneity.

Evaluación de los sesgos de notificación

For all its advantages and strengths, meta‐analysis is a rough statistical approach used to estimate a summary effect size, and it is subject to several limitations. Perhaps most notably, much of the meta‐analysis literature relies on results from publications and other accessible sources, resulting in strong selection bias in a weighted average. In the fields of medicine and psychology, study results displaying negative results or insignificant findings are much less likely to be accepted by scientific journals for publication. As many study results failing to show significance presumably still lie in researchers' file drawers, this publication bias is often referred to as the 'file drawer' problem (Rosenthal 1979). To date, no sufficiently satisfactory solution has been found for correcting this type of bias. If the studies included in a meta‐analysis are not randomly selected, which we believe is generally the case, the distribution of effect sizes tends to be skewed, resulting in a biased weighted average of effect sizes.

In our analyses, we viewed the summary effect size estimate from meta‐analysis as representing the statistic for a sample of studies that tends to be skewed. Statistically, the distribution of effect size estimates for selected studies will likely be truncated, leading to non‐normally distributed data. Because studies not accessible for meta‐analysis are usually those with low or even reversed effect sizes, the summary effect size from meta‐analysis tends to be overestimated. We prepared a funnel plot and examined it for asymmetry.

Síntesis de los datos

We applied the random‐effects meta‐analysis to incorporate heterogeneity across studies. Through this statistical approach, the summary effect size estimate measures the mean of systematically different effects in different studies, while its confidence interval describes uncertainty in the estimate. We calculated between‐study variance to measure such uncertainty; its square root was the estimated standard deviation of study‐specific effects from which the 95% confidence interval can be readily derived.

Análisis de subgrupos e investigación de la heterogeneidad

To investigate heterogenous results and to evaluate whether PCT had different effects based on treatment modality and type we carried out the following subgroup analysis for the primary outcomes. (1) As group and individual TF‐CBT tend to have differential effects, we were interested to see whether PCT performed differently across these modalities. (2) Given that PE and CPT are two of the most common TF‐CBT treatments used, we also explored whether PCT had differential effects based on these specific trauma treatments.

• Treatment modality (individual or group). PCT is administered in both individual and group formats; this variable may affect heterogeneity and treatment outcomes.

• Treatment type (PE or CPT). PCT may have differential efficacy based on the comparator treatment.

Análisis de sensibilidad

We conducted sensitivity analyses for the primary outcomes, focusing on those trials with lowest risk of bias as based on the following criteria: outcome masking, appropriate handling of missing data (ITT; mixed‐model analysis), adequate power, and low levels (< 40%) of post‐randomization treatment loss.