La poliomielitis: representación social y comunicación

A sensitivity analyses was planned to explore the influence of study design factors including sequence generation, allocation sequence concealment, blinding of outcome assessor and presence of intention-to-treat analysis to determine how robust the analyses are. However, it was decided that the planned sensitivity analyses were not appropriate due to the diverse clinical and methodological nature of the studies.

5.4 Results (C)

5.4.1 Description of studies

5.4.1.1 Results of the search

The search strategy identified 22713 citations. Title and abstract screening identified 87 manuscripts and trial records as potentially eligible for this review.

The full text screening excluded 65 manuscripts. See Figure 5-1.

Of the remaining 22 studies, the main contacts for two completed trials identified from clinical trial registers were contacted for further information without any success ^296-298 and one contact person for another completed trial could not be traced^299-300 (see table 5-4Characteristics of studies awaiting assessment); 11 studies are ongoing^301-312 (see table 5-3 Characteristics of ongoing studies) . Of the remaining eight studies 63; 313-320 seven authors were

contacted for further details on study methods and outcome data63; 313-319. Data from one study published in Korean³²⁰ was abstracted by a Korean speaking person. There were insufficient resources to get more information than was available from the published paper. Only one author failed to respond the request for further information³¹⁵.

5.4.1.1.1 Included studies

Eight studies 63; 313-320 including a total of 1007 participants met the inclusion criteria. Detailed information about each is shown in the table 5-1

Characteristics of included studies. Seven studies63; 313-319 were published in English. One study was published in Korean³²⁰.

5.4.1.1.2 Participants

One study targeted carers of stroke survivors with aphasia³¹⁵. The remainder of the studies targeted informal carers of stroke survivors.

5.4.1.1.3 Definition of the index condition for being a current informal carer Five studies 313; 314; 316; 319; 320 did not provide any definition of the index condition for an informal carer and three studies 63; 315; 317; 318 provide vague descriptions.

5.4.1.1.4 Experimental interventions

The studies included in this review are clinically diverse. Therefore it was

decided that broader categories of intervention should be created to reflect the primary focus of the experimental interventions. To ensure that the process of categorisation was unbiased, each review author was asked to independently read an assign an anonymised extract describing the experimental intervention to one of three broad intervention categories. The results were collated and disagreement was resolved by consensus. The three broad types of intervention were:

 Support and information i.e., interventions that provide participants with information to connect them with necessary resources, opportunities or supports.

 Teaching procedural knowledge/ vocational education i.e., interventions that focus on preparing participants for the work of providing care to a stroke survivor and is based on manual or practical activities

 Psycho educational i.e., interventions that reinforce personal strengths, resources and coping skills of participants, in order to for example, to avoid relapse or contribute to their own health and wellness on a long-term basis.

There were six modes of delivery of intervention:

 Face to face

 Telephone

 Group – face to face

 Group – telephone

 Combination of face to face and telephone

 Internet

Four studies tested interventions aimed at providing information and support

316-320, three studies tested psycho education interventions^313-315, one study tested the effects of ‘teaching procedural knowledge’⁶³. Five trials tested experimental interventions delivered face to face 63;313;316-318;320, two trials ^315;319 tested

interventions that were delivered remotely i.e., by telephone ³¹⁵ or Internet³¹⁹. One multi-arm trial³¹⁴ tested a combination of face to face and remote

(telephone) delivery of the intervention. Of the six trials testing face to face experimental interventions, two were delivered to formal groups of participants

313; 316 and the remainder were delivered on an individual basis.

5.4.1.2 Study interventions and comparisons

For details of study interventions and comparisons refer to the Characteristics of included studies table. One trial³¹⁴ compared two alternative forms of

interventions against usual hospital care, comparing social problem solving therapy partnerships versus sham intervention versus usual hospital care. One trial³¹³ used a cross over design in which the participants took part in a psycho education programme, the experimental intervention, in sequence. For the purpose of this review, the end of scheduled follow-up is the end of the first treatment period at 12 weeks.

5.4.1.3 Excluded studies

For details for excluded studies with reasons see table Characteristics of excluded studies.

5.4.1.4 Risk of bias in included studies

For each included trial we extracted information about the method of

randomisation and allocation concealment, blinding of outcome assessment and whether all the randomised patients were accounted for in the analysis.

5.4.1.4.1 Allocation

The inclusion criteria for this analysis required a study to be randomised. Six studies reported an adequately generated allocation sequence 63; 313; 316-320. Two studies63; 317; 318 reported adequately concealed allocation. Two studies did not conceal allocation ^{313; 319}. Four studies were unclear 314; 315; 316; 320.

5.4.1.4.2 Blinding

It is not possible to blind key study personnel or participants in studies which do not use a placebo comparator. However, non-blinding of participants and study personnel is unlikely to introduce bias if the outcome assessment is blinded ³²¹. Four studies report blinding of the outcome assessment/ assessor 63; 313; 314; 317;

318.

5.4.1.4.3 Incomplete outcome data

Missing continuous outcome data due to attrition is an issue across all studies.

However, the extent and nature of attrition varies across trials. For details of the amount and distribution of missing outcome data, the reasons provided for missing outcome data, the investigators handling of missing data as well as the clinical context and judgement on risk of bias see table Risk of bias in included studies. Missing outcome data is likely to be associated with the outcome or perceived relevance of the intervention to the study participants.

5.4.1.4.4 Selective reporting

The majority of included studies appear to be free from suggestion of selective outcome reporting.

5.4.1.4.5 Other potential sources of bias

Two studies did not provide information on baseline characteristics ^{63; 314}, there was baseline imbalance in two studies 315; 317; 318 and insufficient information in one study to permit judgement about baseline imbalance³²⁰.