Woodfordetal.SystematicReviews2013,2:9 http://www.systematicreviewsjournal.com/content/2/1/9 PROTOCOL Open Access Psychological treatments for common mental health problems experienced by informal carers of adults with chronic physical health conditions (Protocol) Joanne Woodford1*, Paul Farrand1, David Richards2 and David J Llewellyn2 Abstract Background: Improved life expectancy is resulting inincreased outpatient treatmentof people with chronic physical health conditions and reliance onthe provision of informal care in the community. However, informal care is also associated with increased risk of experiencingcommon mental health difficulties such as depression and anxiety. Currently there is a lack ofevidence-based treatments for such difficulties, resulting in poor health outcomes for both theinformalcarer and care recipient. Methods/Design:Electronic databases will be systemically searched for randomised controlled trials examining the effectiveness of psychological interventions targeted attreatingdepressionor anxiety experienced by informal carers of patients with chronic physical health conditions. Database searches will be supplementedby contact with experts, reference and citationchecking and grey literature.Both publishedand unpublished research in English language will be reviewed with nolimitationson year or source. Individual, group and patient-carer dyad focused interventions will be eligible.Primary outcomes ofinterest will be validated self-report or clinician administered measures of depressionor anxiety.If data allows a meta-analysis will examine: (1)the overall effectiveness of psychological interventions in relation to outcomesofdepressionor anxiety; (2) intervention components associated with effectiveness. Discussion: Thisreview will provide evidence on theeffectivenessof psychologicalinterventions for depression and anxiety experienced byinformalcarers ofpatientswith chronic physical health conditions. Inaddition, it will examine interventioncomponents associated with effectiveness. Results will inform thedesign and development of a psychological intervention for carers of people with chronic physical health conditions experiencingdepression and anxiety. PROSPERO registration number: CRD42012003114 Keywords: Caregivers, Chronic physical health condition, Depression,Anxiety, Treatment,Systematic review *Correspondence:[email protected] 1MoodDisordersCentre,Psychology,CollegeofLifeandEnvironmental Sciences,UniversityofExeter,Exeter,UK Fulllistofauthorinformationisavailableattheendofthearticle ©2013Woodfordetal.;licenseeBioMedCentralLtd.ThisisanOpenAccessarticledistributedunderthetermsofthe CreativeCommonsAttributionLicense(http://creativecommons.org/licenses/by/2.0),whichpermitsunrestricteduse, distribution,andreproductioninanymedium,providedtheoriginalworkisproperlycited. Woodfordetal.SystematicReviews2013,2:9 Page2of7 http://www.systematicreviewsjournal.com/content/2/1/9 Background A number of meta-analyses have been undertaken to Advances in public health and medical technology have identify factors associated with positive outcomes in in- resulted in continued increases in life expectancy across formal carers, such as caregiver burden, knowledge, de- developed countries [1]. Within the UK alone the pression and symptoms of care recipients. Commonly projected rise in adults aged over 80 years is due to rise such meta-analyses have included a large number of po- from 2.9 million adults in 2010 to 5.9 million in 2035 tential factors, such as respite and day care, knowledge [2]. Theseincreases inlife expectancy are presenting sig- and training, group- and individual-based interventions, nificant challenges to existing healthcare systems with type of setting and various caregiver characteristics [24]. regards to the management and treatment of patients Additionally they have focused upon a variety of specific with chronic or disabling illnesses [1,3]. This is mani- patient carer groups such as stroke [25], cancer [26] and festing itself in an increased reliance on informal carers dementia[23,27].Farlessattentionhashoweverbeendi- as a fundamental part of patient management which rected towards identifying specific psychological in- has become important following an increasing emphasis terventions that may be targeted at the treatment of upon outpatient treatment of patients with chronic phy- depression and anxiety. Where this focus has been in- sical health conditions [4]. Currently around 5 million cluded within the meta-analysis as treatment modera- people in the UK provide informal care to someone with tors, psychological treatments mostly consistent with aphysical ormental healthdifficulty[5]. elements of cognitive behavioural therapy (CBT) have Theshift tooutpatient treatmentalongside aconcomi- been identified to have a small effect on caregiver de- tant increase in the role of informal carer in patient pression [24,28]. Furthermore, one review has identi- management and treatment has led to a reduction in pa- fied cognitive reframing, a specific component commonly tient hospital and physician care as well as delaying the found within CBT, as an intervention component promi- receipt of nursing home care [6]. However given de- sing for depressed informal carers of dementia patients mands associated with supporting the treatment and re- [21]. covery of patients with a physical or mental health To date little research has sought to identify specific difficultynowbeingplaceduponinformalcarers,thepo- intervention components that have been utilised in in- tential is that costs are simply being shifted elsewhere. terventions targeting depression and anxiety in carers of Informal care is not only associated with greater risks adults with chronic physicalhealth conditions. Addition- of poor mental and physical health [7-10] but addi- ally, when such components have been identified tional personal and societal costs arising from reduc- through approaches such as systematic reviews, it has tions in hours of paid work, restriction in social and been reportedthattoolittle attention hasthenbeenpaid recreational activities [11], and sleep disturbances [12]. to identifying the specific components associated with Additionally poor mental health in carers may also ne- effectiveness [29,30]. Recent systematic reviews have gatively impact on outcomes associated with the care re- therefore examined not only the overall effectiveness of cipient[13]. interventions but also the specific intervention compo- A clear need therefore exists to develop evidence- nents associated with their effectiveness [30,31]. This based psychological interventions to support the long- systematic review therefore seeks to examine both the term emotional needs of informal carers. However such overall effectiveness of psychological interventions for long-term emotional needs of carers have been neglec- depressed or anxious carers and specific intervention ted across a range of chronic physical health conditions components associated with effectiveness. The identi- [14-18]. Furthermore, services that do exist to provide fication of effective intervention components utilised emotional support are often inadequately developed and in interventions targeting depression and anxiety in are generally not tailored to address the unique difficul- carers of adults with chronic physical health conditions is ties carers’ experience [19]. Such difficulties may include an important next step to inform the future design and the management of behavioural problems [7], physical developmentofevidence-basedtreatments. impairments [20], cognitive decline [21] and the deve- lopment of communication techniques [22]. Developing Objectives interventions have however potentially been hindered First, to undertake a comprehensive systematic review given that the needs of carers often change dependent andmeta-analysisexaminingtheeffectivenessofpsycho- upon the course of the chronic physical health condition logical interventions targeted at treating emotional diffi- of the care recipient, the setting care is provided in and culties, such as depression or anxiety, across a range of length of time care has been provided [17]. Recognition carer-care recipient populations. Second, to identify of the unique and multifaceted needs of informal carers intervention components associated with effectiveness. has led to the suggestion that multicomponent interven- The results of the systematic review will also be used to tionsarerequired[17,23]. feed into thedevelopmentofanevidence-basedcomplex Woodfordetal.SystematicReviews2013,2:9 Page3of7 http://www.systematicreviewsjournal.com/content/2/1/9 intervention for carers of people with chronic physical Outcomes health conditions using the Medical Research Council’s Studies eligible for inclusion will have a primary or sec- (MRC)guidance [32,33]. ondary outcome measurement of a validated self-report or clinician administered measure of depression or anx- iety that elicits continuous data. Outcomes of caregiver Methods burden and quality of life will also be examined. Drop- The review will follow the Centre for Reviews and Dis- out rates will also be recorded. Outcomes for any time semination (CRD) guidance on undertaking systemic re- period will be eligible for inclusion. However, in the case views [34] and be reported to established criteria [35]. of studies reporting multiple time points the follow-up The review is registered with the PROSPERO Inter- time point used for analysis will be the longest time national Prospective Register of Systematic Reviews point ≤6months. (registrationnumberCRD42012003114). Studydesign Inclusionandexclusioncriteria Only randomised controlled trials using a method of Population randomsequence generation andallocation concealment Eligible populations are informal adult (aged 16 years assessed as low or unclear risk of bias using the and older) carers of adults with chronic physical health Cochrane Collaboration’s Risk of Bias tool [38] will be conditions who are experiencing depression or anxiety. includedwithin thereview. Formal diagnosis of depression or anxiety will not be re- quired. No limitations will be placed on severity of de- pression or anxiety (though it is estimated carers will be Searchstrategy depressed or anxious), length of time caring, chronic A comprehensive search will be conducted on the fol- physicalhealthconditionofthepersoncaredfor,orrela- lowing electronic databases: Cumulative Index to Nurs- tionship to person cared for. Informal carers will be de- ing and Allied Health Literature (CINAHL); Excerpta fined as non-professionals who support people who are Medica DataBase (EMBASE); PsychInfo; Medline; Social sick, infirm or disabled [36]. Commonly this group is Science Citation Index; Applied Social Sciences Index made up of the patients’ close family, however non- and Abstracts (ASSIA) and the Cochrane Central Regis- family informal carers will be eligible for inclusion. ter ofControlled Trials (CENTRAL). Reference lists and Given the recognition that provision of care is dynamic citations will be hand searched for all included stud- and fluctuates from providing intensive assistance on a ies to identify further studies. The results of the data- daily basis to more infrequent support no constraints base searches will be analysed to identify journals that will be placed on how much assistance informal carers contain the largest number of included studies which provide[37]. will be hand searched for recent publications and con- ference abstracts (less than 12 months). Trial registers Interventions www.ClinicalTrials.gov and www.who.int/trialsearch/ will Thereview will include psychological or psychosocial in- also be searched to identify on-going or unpublished terventions that are targeted at depression or anxiety. trials. Experts in the field will be contacted to further There will be no limitation in terms of psychological identify unpublished or ongoing trials. An information theory informing the intervention, the person delivering specialist was consulted to build the search strategy the intervention or the setting in which the intervention using medical subject headings (MeSH). The Ovid is delivered. Group, one-to-one and unsupported inter- MEDLINE search strategy can be found in Additional ventions will be included. Interventions for the carer- file 1. patient dyad will also be included as long as a target of theintervention iscarerdepressionoranxiety. Studyselection All titles and abstracts will be screened by JW and a sec- Comparators ond researcher. Full paper review to determine inclusion Only interventions compared with an inactive control will be conducted independently by JW and PF. Cohen’s will be considered. This may include: a waiting list con- Kappa will be calculated to determine agreement in trol; treatment-as-usual (normally defined as standard selecting studies in accordance with the exclusion / in- care provided by a general practitioner/family doctor); clusion criteria. Any discrepancies will be resolved by no treatment and attention-controls. Interventions com- discussion and, if consensus cannot be reached, a third pared with another active intervention will not be eli- member of the research team will make the final gible forinclusion. decision. Woodfordetal.SystematicReviews2013,2:9 Page4of7 http://www.systematicreviewsjournal.com/content/2/1/9 Dataextraction Methodologicalquality Data extraction will be conducted by JW and a re- The Cochrane Collaboration’s Risk of Bias tool [38] will searcher not associated with the research team. Discrep- be adopted to appraise the methodological quality of the ancies will be discussed and if consensus is not reached included studies. This will be undertaken independently discussion will be held with PF. A data extraction form by JW and a reviewer not associated with the research specifically for this review has been developed upon team. Ratings will be compared and any discrepancies guidancefrom theCRD[34]. discussed, and if consensus is not reached, further dis- To meet the second objective of the review there will cussion will be held with PF. The tool will examine risk also be a specific focus on extracting information re- of selection, performance, attrition and reporting bias. lating to intervention components and patient cha- To detect reporting bias attempts will be made to obtain racteristics in addition to the standard extraction of study protocols for all included studies either via pub- information (for example, identification features, study lishedprotocols,trialdatabasesoremailingthestudyau- characteristics, primary outcome measurements, statis- thors. Comparisons will be made with the outcome tical approaches and primary results). Intervention com- measurements reported in the protocol and the paper. ponents extracted from the data are partially based upon In addition, outcomes reported in the methods section those used in a previous review examining intervention will be compared with outcomes reported in the results components associated with increased effectiveness in section. In the event of discrepancies study authors will diet and physical activity interventions [31]. Specific- be contacted to identify potential reasons, such as ally the following will be extracted: (1) theoretical changes to the study protocol, or to request missing framework (for example, cognitive therapy, behaviour data. In addition, the quality of primary outcome mea- therapy, interpersonal therapy, psychodynamic therapy); sures and whether a power calculation was conducted (2) behaviour change techniques (for example, problem will be assessed. The quality of outcome measurements solving, goal setting, relapse prevention) based on a used will be examined in terms of reliability through in- taxonomy of 137 behaviour change techniques [39]; (3) ternal consistency and test-retest reliability [40]. Only mode of delivery (for example, individual face-to-face, studies using outcome measurements of at least ac- telephone, email, group, unsupported self-help); (4) ceptable internal consistency and test-retest reliability group size for group-based interventions; (5) clinician (Cronbach’salpha≥0.70)willbeincluded[40].Allfindings delivering treatment (for example, nurse, general prac- will be summarised within a table to allow easy compari- titioners, clinical psychologist); (6) training received by sonacrossstudies. the clinicians delivering the treatment; (7) treatment intensity (for example, duration of treatment, number Datasynthesisandanalysis of sessions, length of sessions); (8) whether the treat- Effectsizeestimates ment is manualised (yes or no); (9) measurement of If possible with available data, a meta-analysis will be treatment integrity (yes or no); and (10) treatment setting conducted using Comprehensive Meta-Analysis Version (for example, primary care, secondary care). In addition, 2.0 [41]. Post-treatment between group standardised specificcharacteristicswillbeextracted for boththe carer mean difference effect size will be calculated using (for example, age, ethnicity, severity of depression or Hedges’ g from the outcomes relating to depression, anxiety at baseline, length of time caring, relationship anxiety, quality of life and caregiver burden separately. to person cared for and receipt of formal care in the Where multiple time points are reported the longest home) and the adult with the chronic physical health follow-up time point will be taken ≤6 months. Means condition (for example, age, chronic physical health and standard deviations of post-outcome measurement condition, severity of chronic physical health condition scores will be requested from authors if not reported andmentalhealthandotherchronicphysicalhealthco- within the paper. Heterogeneity is expected and there- morbidities). The data extraction form can be found in fore a random-effect model will be used. In the event Additionalfile2. that there is no evidence of heterogeneity between stud- Intervention components will be extracted from ies a fixed-effect model will be selected. The presence of published papers however all authors will also be statistically significant heterogeneity will be examined contacted to obtain trial protocols and treatment using Cochrane’s test of heterogeneity (Q statistic) and manuals associated with the delivery of the interv- the I2 statistic will also be reported to quantify the de- ention to enable more detailed coding to take place. gree of heterogeneity [42,43]. I2 values of heterogeneity Interventions will be coded by JW and PF indepen- will be considered low, moderate or high using cutoffs dently and discrepancies will be resolved through dis- of 25%, 50% and 75%, respectively [43]. If intention-to- cussion and if required a third researcher will be treat data are available these will be used to calculate ef- consulted. fect sizes, with completer used when intention-to-treat Woodfordetal.SystematicReviews2013,2:9 Page5of7 http://www.systematicreviewsjournal.com/content/2/1/9 data are unavailable. With studies that compare two heterogeneity.Consistentwithothermeta-analyses[51,52] treatment conditions that are eligible for inclusion, com- subgroup analyses will be considered statistically signifi- parisons will be analysed separately with the sample size cant if a P value of ≤0.10 is obtained. In the event that within the control condition halved. Comparisons will thereisnotenoughinformationinrelationtocomponents be analysed separately with the sample size within the ofinterventionstosupportameta-analysis[30]anarrative intervention arm halved when two control conditions synthesiswillbeundertakentosummarisethesefindings. areincluded. Discussion Funnelasymmetry This review will examine the effectiveness of psycho- Egger’sTestoftheIntercept[44]willbeusedtoexamine logical interventions for informal carers of people with funnelplotasymmetrytoinvestigatepossible publication chronic physical health conditions experiencing depres- bias and other potential sources of asymmetry (for ex- sion or anxiety. Currently there is no comprehensive re- ample, language bias, potential inclusion of small studies view of psychological interventions for informal cares with poor methodological rigour, heterogeneity) [44]. that also systematically examines both the quality of Egger’s Test of the Intercept will only be conducted if a available evidence and intervention components asso- minimum of 10 studies are included within the meta- ciated with effectiveness. Thereby this review seeks to analysis [45]. The trim and fill procedure [46] will be both examine gaps in the evidence base for future re- used to calculate an effect size taking into account po- search and also to map intervention components asso- tentialpublication bias. ciated with effectiveness. The identification of specific intervention components associated with effectiveness Sensitivityanalysis willaidthetranslationoftheexistingevidencetothede- Sensitivity analyses will be undertaken to examine the velopment of new interventions optimising these com- extent to which results obtained may be influenced by ponents. Thus, the mapping of such components is a the selective reporting of outcomes. The maximum bias first step towards developing a psychological treatment bound approach [47-50] will be adopted with new treat- for informal carers that maximises the use of behav- ment effect and confidence intervals calculated by ioural change techniques and delivery factors associated adding the bias bound value to the original pooled effect with effectiveness in order to meet objectives within estimate to examine the robustness of findings [48]. Fur- Phase I of the MRC’s guidance [32,33] for developing ther sensitivity analysis will also be conducted by tem- complex interventions. porarily dropping from the analysis: small studies (n ≤20); unpublishedstudies;studieswithhighattrition(≥30%);and Additional files studies where outcome measurements of depression and anxietyarereportedasprimaryorsecondaryoutcomemea- Additionalfile1:OvidMEDLINESearchStrategy. surementstoexaminewhetherresultsremainconsistent. Additionalfile2:DataExtractionForm. Moderatoranalysis Abbreviations When number of studies addressing particular mode- ASSIA:AppliedSocialSciencesIndexandAbstracts;CBT:Cognitive rators permit, moderator analysis will be undertaken to behaviouraltherapy;CENTRAL:CochraneCentralRegisterofControlledTrials; CINAHL:CumulativeIndextoNursingandAlliedHealthLiterature; examine intervention components, methodological and CRD:CentreforReviewsandDissemination;EMBASE:ExcerptaMedica participant characteristics of studies associated with DataBase;MeSH:MedicalSubjectHeadings;MRC:MedicalResearchCouncil; effectiveness.Specificallythefollowingmoderatorswill PROSPERO:Internationalprospectiveresisterofsystematicreviews. be examined: (1) chronic physical health condition of the Competinginterests carerecipient;(2)theoreticalframework(forexample,cog- Theauthorsdeclarethattheyhavenocompetinginterests. nitive therapy, behaviour therapy); (3) behaviour change techniques used (for example, problem solving, goal Authors’contributions JWconceivedanddesignedthestudyprotocolandwrotethemanuscript. setting, relapse prevention); (4) mode of delivery (for PFcontributedtotheconception,designandhadsubstantialinvolvement example, individual face-to-face, telephone, group); (5) inthedraftingofthemanuscript.DRandDLmadecontributionstothe duration of treatment; (6) number of treatment sessions; design,criticalevaluationofintellectualcontentandassistedwithdrafting themanuscript.Allauthorshaveapprovedthefinalmanuscript. (7) baseline severity of depression or anxiety; (8) diagnosis ofdepressionoranxiety(yesorno);(9)recruitmentsetting. Acknowledgements Moderators will be examined through subgroup analysis ThispaperissupportedbytheDunhillMedicalTrustintheformofa with standardised mean difference effect sizes calculated doctoralresearchtrainingfellowshipawardedtothefirstauthor(JW).The using Hedges’g statistic using a random-effects model. DunhillMedicalTrustisamemberoftheAssociationofMedicalResearch CharitiesAMRCandaNationalInstituteforHealthResearch(NIHR) Q and I2 statistics will also be reported as a measure of recognisednon-commercialpartner. Woodfordetal.SystematicReviews2013,2:9 Page6of7 http://www.systematicreviewsjournal.com/content/2/1/9 Authordetails 24. SörensenS,PinquartM,DubersteinP:Howeffectiveareinterventionswith 1MoodDisordersCentre,Psychology,CollegeofLifeandEnvironmental caregivers?Anupdatedmeta-analysis.Gerontologist2002,42:356–372. Sciences,UniversityofExeter,Exeter,UK.2UniversityofExeterMedical 25. LeggLA,QuinnTJ,MahmoodF,WeirCJ,TierneyJ,StottDJ,SmithLN, School,Exeter,UK. LanghorneP:Non-pharmacologicalinterventionsforcaregiversofstroke survivors(review).CochraneDatabaseSystRev2011,10:CD008179. Received:17October2012Accepted:15January2013 26. HopkinsonJB,BrownJC,OkamotoI,Addington-HallJM:Theeffectiveness Published:31January2013 ofpatient-familycarer(couple)interventionsforthemanagementof symptomsandotherhealth-relatedproblemsinpeopleaffectedby cancer:asystematicliteraturesearchandnarrativereview.JPain References SymptomManage2012,43:111–142. 1. ChristensenK,DoblhammerG,RauR,VaupelJW:Ageingpopulations:the 27. VanMierloLD,MeilandFJ,VanderRoestHS,DröesRM:Personalised challengesahead.Lancet2009,374:1196–1208. caregiversupport:effectivenessofpsychosocialinterventionsin 2. OfficeofNationalStatistics:AnExecutivesummary,2010-basedNPPReference subgroupsofcaregiversofpeoplewithdementia.IntJGeriatrPsychiatry Volume.London:TheStationeryOffice;2012. 2012,27:1–14. 3. LubitzJ,CaiL,KramarowE,LentznerH:Health,lifeexpectancy,andhealth 28. PinquartM,SörensenS:Helpingcaregiversofpersonswithdementia:which carespendingamongtheelderly.NewEnglJMed2003,349:1048–1055. interventionsworkandhowlargearetheireffects?IntPsychogeriatr2006, 4. WakefieldBJ,HayesJ,BorenSA,PakY,DavisJW:Strainandsatisfactionin 18:577–595. caregiversofveteranswithchronicillness.ResNursHealth2012,35:55–69. 29. CaneJ,O’ConnorD,MichieS:Validationofthetheoreticaldomains 5. NHSInformationCentre:SurveyofCarersinhouseholdsinEngland2009/10. frameworkforuseinbehaviourchangeandimplantationresearch. Leeds:TheNHSInformationCentreforHealthandSocialCare;2010. ImplementSci2012,7:37. 6. VanHoutvenCH,NortonEC:Informalcareandhealthcareuseofolder adults.JHealthEcon2004,23:1159–1180. 30. LorencattoF,WestR,MichieS:Specifyingevidence-basedbehavior changetechniquestoaidsmokingcessationinpregnancy.NicotineTob 7. PinquartM,SörensenS:Differencesbetweencaregiversand Res2012,14:1019–1026. noncaregiversinpsychologicalhealthandphysicalhealth:ameta- analysis.PsycholAging2003,18:250–267. 31. GreavesCJ,SheppardKE,AbrahamC,HardemanW,RodenM,EvansPH, SchwarzP,IMAGEStudyGroup:Systematicreviewofreviewsof 8. CoeNB,vanHoutvenCH:Caringformomandneglectingyourself?The interventioncomponentsassociationwithincreasedeffectivenessin healtheffectsofcaringforanelderlyparent.HealthEcon2009, 18:991–1010. dietaryandphysicalactivityinterventions.BMCPublicHealth2011, 11:119. 9. ShawWS,PattersonTL,SempleSJ,HoS,IrwinMR,HaugerRL,GrantI: 32. CampbellM,FitzpatrickR,HainesA,KinmonthAL,SandercockP, Longitudinalanalysisofmultipleindicatorsofhealthdeclineamong spousalcaregivers.AnnBehavMed1997,19:101–109. SpiegelhaterD,TyrerP:Frameworkfordesignandevaluationofcomplex interventionstoimprovehealth.BMJ2000,321:694–696. 10. CanniscioCC,JonesC,KawachiI,ColditzGA,BerkmanL,RimmE: 33. CraigP,DieppeP,MacintyreS,MichieS,NazarethI,PettcrewM: Reverberationsoffamilyillness:alongitudinalassessmentofinformal caregivingandmentalhealthstatusintheNurses’HealthStudy.AmJ Developmentandevaluatingcomplexinterventions:thenewMedical PublicHealth2002,92:1305–1311. ResearchCouncilguidance.BMJ2008,337:a1655. 34. CentreforReviewsandDissemination(CRD):SystematicReviews:CRD’s 11. SalvaJ,AlmeidaDM,DaveyA,ZaritSH:Routineassistancetoparents: GuidanceforUndertakingReviewsinHealthCare.3rdedition.York:University effectsondailymoodandotherstressors.JGerontolBPsycholSciSocSci 2008,63:S154–S161. ofYork;2009. 35. MoherD,LiberatiA,TetzlaffJ,AltmanDG,ThePRISMAGroup:Preferred 12. PhillipsAC,GallagerS,HuntK,DerG,CarrollD:Symptomsofdepressionin ReportingItemsforSystematicReviewsandMeta-Analyses:ThePRISMA non-routinecaregivers:theroleofcaregiverstrainandburden.BrJClin Psychol2009,48:335–346. Statement.PLoSMed2009,6:e1000097. 36. SingletonN,AyeMaungN,CowieA,SparksJ,BumpsteadR,MeltzerH: 13. PerrinPB,HeesackerM,StidhamBS,RittmanMR,Gonzalez-RothiLJ: Mentalhealthofcarers.London:TheStationeryOffice;2002. Structuralequationmodelingoftherelationshipbetweencaregiver psychosocialvariablesandfunctioningofindividualswithstroke.Rehabil 37. SzinovaczME,DaveyA:Changesinadultchildcaregivernetworks. Psychol2008,53:54–62. Gerontologist2007,47:280–295. 14. MurrayJ,YoungJ,ForsterA,AshworthR:Developingaprimarycare- 38. HigginsJ,AltmanDG:Assessingriskofbiasinincludedstudies.In basedstrokemodel:theprevalenceoflonger-termproblems CochraneHandbookforSystematicReviewsofInterventions.Editedby experiencedbypatientsandcarers.BrJGenPract2003,53:803–807. HigginsJP,GreenS.Chichester:JohnWiley&Sons;2008:187–241. 15. SimonC,KumarS,KendrickT:Formalsupportofstrokesurvivorsand 39. MichieS,JohnstonM,FrancisJ,HardemanW,EcclesM:Fromtheoryto theirinformalcarersinthecommunity:acohortstudy.HealthSocCare intervention:mappingtheoreticallyderivedbehavioural Community2008,16:582–592. determinantstobehaviouralchangetechniques.ApplPsychol2008, 16. SimonC,KumarS,KendrickT:Cohortstudyofinformalcarersoffirst-time 57:660–680. strokesurvivors:profileofhealthandsocialchangesinthefirstyearof 40. FitzpatrickR,DaveyC,BuxtonMJ,JonesDR:Evaluatingpatient-based caregiving.SocSciMed2009,699:404–410. outcomemeasuresforuseinclinicaltrials.HTA1998,2:14. 17. WilkinsonAM,LynnJ:Caregivingforadvancedchronicillnesspatients. 41. BorensteinM,HedgesL,HigginsJ,RothsteinH:ComprehensiveMeta-analysis TechRegAnesthPainManag2005,9:122–132. Version2.Englewood,NJ:Biostat;2005. 18. ArkseyH,HirstM:Unpaidcarers’accesstoanduseofprimarycare 42. HigginsJP,ThompsonSG:Quantifyingheterogeneityinmeta-analysis. services.PrimHealthCareResDev2005,6:101–116. StatMed2002,21:1539–1558. 19. LundhU:Familycarers,4:Designingservicestosupportfamilycarersin 43. HigginsJP,ThompsonSG,DeeksJJ,AltmanDG:Measuringinconsistency Sweden.BrJNurs1999,8:787–790. inmeta-analyses.BMJ2003,327:557–560. 20. PinquartM,SörensenS:Associationsofstressorsandupliftsofcaregiving 44. EggerM,DaveySmithG,SchneiderM,MinderC:Biasinmeta-analysis withcaregiverburdenanddepressivemood:ameta-analysis.JGerontol detectedbyasimple,graphicaltest.BMJ1997,315:629–634. BPsycholSciSocSci2003,58:112–128. 45. SterneJAC,GavaghanD,EggerM:Publicationandrelatedbiasinmeta- 21. Vernooij-DassenM,DraskovicI,McCleeryJ,DownsM:Cognitivereframing analysis:Powerofstatisticaltestsandprevalenceintheliterature.JClin forcarersofpeoplewithdementia.CochraneDatabaseSystRev2011, Epidemiol2000,53:1119–1129. 11:CD005318. 46. DuvalS,TweedieR:Trimandfill:Asimplefunnel-plot–basedmethodof 22. RoslandAM,PietteJD:Emergingmodelsformobilizingfamilysupportfor testingandadjustingforpublicationbiasinmeta-analysis.Biometrics chronicdiseasemanagement:astructuredreview.ChronicIlln2010, 2000,56:455–463. 6:7–21. 47. CopasJ,JacksonD:Aboundforpublicationbiasbasedonthefractionof 23. BrodatyH,GreenA,KoscheraA:Meta-analysisofpsychosocial unpublishedstudies.Biometrics2004,60:146–153. interventionsforcaregiversofpeoplewithdementia.JAmGeriatrSoc 48. WilliamsonPR,GambleC:Applicationandinvestigationofaboundfor 2003,51:657–664. outcomereportingbias.Trials2007,8:9. Woodfordetal.SystematicReviews2013,2:9 Page7of7 http://www.systematicreviewsjournal.com/content/2/1/9 49. KirkhamJJ,DwanKM,AltmanDG,GambleC,DoddS,SmythR,Williamson PR:Theimpactofoutcomereportingbiasinrandomisedcontrolledtrials onacohortofsystematicreviews.BMJ2010,40:c365. 50. DwanK,GambleC,Kolamunnage-DonaR,MohammedS,PowellC, WilliamsonPR:Assessingthepotentialforoutcomereportingbiasina review:atutorial.Trials2010,11:52. 51. HesserH,WeiseC,ZetterqvistWestinV,AnderssonG:Asystematicreview andmeta-analysisofrandomizedcontrolledtrialsofcognitive- behavioraltherapyfortinnitusdistress.JClinPsychol2011,31:545–553. 52. FarrandP,WoodfordJ:Impactofsupportoneffectivenessofwritten behaviouralself-help:Asystematicreviewandmeta-analysisof randomisedcontrolledtrials.JClinPsychol2013,33:182–195. doi:10.1186/2046-4053-2-9 Citethisarticleas:Woodfordetal.:Psychologicaltreatmentsfor commonmentalhealthproblemsexperiencedbyinformalcarersof adultswithchronicphysicalhealthconditions(Protocol).Systematic Reviews20132:9. Submit your next manuscript to BioMed Central and take full advantage of: • Convenient online submission • Thorough peer review • No space constraints or color figure charges • Immediate publication on acceptance • Inclusion in PubMed, CAS, Scopus and Google Scholar • Research which is freely available for redistribution Submit your manuscript at www.biomedcentral.com/submit