Ulighed i deltagelse ved kolorektal cancerscreening i Region Nordjylland

Abstract

Baggrund: Screeningsprogrammet blev indført i Danmark i 2014, og består af en indledende screening ved indsendelse af afføringsprøve, samt en opfølgende koloskopi hvis der findes blod. Dette speciale omhandler ulighed i deltagelse i indledende del af screeningen. Ulighed i deltagelse kan betyde ulighed i screeningsprogrammets effekt for nogle befolkningsgrupper. Problemformuleringen søger derfor svar på, hvordan udformning af udviklingstiltag kan mindske ulighed i deltagelse og øge den overordnede deltagelse i indledende screening, samt hvorledes disse kan implementeres og evalueres. Metode: Den metodiske ramme for dette speciale var mixed methods. I mixed methods-undersøgelsen indgik tre delundersøgelser; statistisk analyse, dokumentanalyse og semistrukturerede interviews. Udviklingstiltagene blev fremanalyseret med udgangspunkt i én samlet analyse af de tre delundersøgelser. Ved logistiske regressionsanalyser estimeredes risikoen for ikke at deltage i screening på baggrund af sociodemografiske variable. På baggrund af dette udpegedes målgruppen for tiltagene. Syv semistrukturerede interviews af deltagere i screeningsprogrammet og en dokumentanalyse bidrog til udvikling af indholdet i tiltagene. Dette ved fortolkning med Health Belief Model som teoretisk referenceramme for interviews og med inddragelse af health literacy ved dokumentanalyse. Leavitts systemmodel blev anvendt til forandringsanalyse af udviklingstiltagene, med henblik på at afdække organisatoriske påvirkninger og mulige konsekvenser ved implementering. Resultater: Ud af 93.500 inviterede borgere i 2014 og 2015 deltog 62.995 (67,37%) i den kolorektale cancerscreening. Laveste deltagelse forekom for enlige (54.2%). Multivariat logistisk regressionsanalyse viste, at enlige var i øget risiko for ikke at deltage (OR 1,63 KI95% 1,56;1,70), sammenlignet med samlevende efter justering for køn, alder, uddannelsesniveau, indkomstkvartiler og immigrationsstatus. Semistrukturerede interviews frembragte 6 kategorier: beslutning truffet på forhånd, forholdet til afsender, praktiske forhold, anvendelse af invitationsmaterialet, risikotanker og omgangskreds. Efter analyse viste fortolkning ved Health Belief Model, at grunden til ikke-deltagelse kan være manglende signaler til handling og en lav grad af refleksion. I dokumentanalysen vurderedes det, at dokumenterne ikke supplerer hinanden godt som samlet beslutningsværktøj, og at der skal foretages ændringer for at tilpasse dette. Den samlede mixed methods-analyse udmundede i fire udviklingstiltag: 1) udsendelse af forhåndsmeddelelse, 2) geografisk inddelte invitationsområder, 3) samtale med egen læge og 4) revideret invitationsbrev. Forandringsanalyse viste ændringer indenfor teknologi, struktur, aktører, opgaver og omgivelser. Evaluering kunne foretages ved procesevaluering, hvor de identificerede elementer i Leavitts systemmodel kan fungere som udgangspunkt for både proces- og evalueringsmål. Konklusion: I Region Nordjylland ses en ulighed i deltagelse mellem enlige og samlevende, idet enlige har 1,63 gange risikoen for ikke-deltagelse, sammenlignet med samlevende. Dermed opnår enlige ikke screeningsprogrammets gavnlige effekt i samme grad. Ved at give enlige flere signaler til handling og skabe øget refleksion kan deltagelsen muligvis øges og uligheden mindskes. Fire udviklingstiltag foreslås for at opnå dette; udsendelse af en forhåndsmeddelelse, revideret invitationsbrev, invitation efter geografiske områder og samtale med egen læge. Det vurderes at udviklingstiltagene vil medføre forandringer i organisationen for tarmkræftscreeningen. Effekten af tiltagene skal løbende evalueres samtidig med implementering.

Background: Colorectal cancer screening was introduced in Denmark in 2014 and included an initial screening, testing for blood in a stool sample, and a following colonoscopy if blood were found. This master thesis revolved around the initial screening. Inequality in participation by sociodemographic predictors, may lead to inequalities in the benefits of participation for some subgroups. The problem statement therefore asked, how the development of interventions can reduce inequalities in participation and increase overall uptake of initial screening procedure, as well as how these interventions can be implemented and evaluated. Methods: The methodological framework for this thesis is mixed methods. In the mixed methods study, three substudies are included; statistical analysis, document analysis and semi-structured interviews. Interventions were developed, using a combined analysis of the results from the substudies. Using logistic regression analyses, the risk of non-participation, were estimated for each sociodemographic variable. From this a target population was found. Seven semi-structured interviews with participants and a document analysis contributed to the substance of the interventions. This was done by interpreting the results, using the Health Belief Model as theoretical framework. Leavitt’s diamond was applied, in an analysis of change, to uncover organisational changes and consequences of intervention implementation. Results: From 93,500 invited citizens, 62,995 (67.37%) participated in colorectal cancer screening. Single individuals had the lowest participation proportion (54.2%). Logistic regression analyses showed that singles had an increased risk of non-participation (OR 1.63 CI95% 1.56;1.70), compared to individuals with a registered partner, after adjustment for gender, age, educational level, income and immigration status. Semi-structured interviews resulted in six categories: Decision made previous to invitation, relation to sender, practical circumstances, use of the invitation material, thoughts of risk, and social circle. Interpretation of interviews, using the Health Belief Model, showed that non-participation could be due to a lack of cues to action and limited reflections. The document analysis assessed that the documents did not supplement each other well as a combined decision aid and alterations are therefore needed. The overall mixed methods analysis resulted in four intervention strategies: 1) distribution of an advance notification letter, 2) inviting geographical areas simultaneously, 3) general practitioner involvement and 4) revised invitation letter. The analysis of change resulted in changes in technology, structure, people, tasks and surroundings. Evaluation could be conducted using process evaluation, in which the identified elements from Leavitt’s diamond could be incorporated as evaluator measures, for both process- and implementation goals. Conclusion: Inequality in participation was evident in the region of Northern Jutland, Denmark, as singles had a 1.63-fold risk of non-participation, compared to individuals with a registered partner. Thereby singles do not obtain the benefits of screening, to the same extend. By giving singles additional cues to action and enhancing reflections, the overall uptake may be increased and inequality decreased. Four interventions were suggested to accomplish this. These interventions will cause organisational changes. The effects of the interventions must be evaluated continuously while implementation is occurring.

A master thesis from Aalborg University

Ulighed i deltagelse ved kolorektal cancerscreening i Region Nordjylland

[Inequality in participipation for colorectal cancer screening in the Region of Northern Jutland, Denmark]