Factors influencing participation in a cancer mortality followback survey: Palliative Medicine

Gomes B., Natalia Calanzani, Hall S., Koffman J., McCrone P., Higginson I.J.

Research output: Contribution to conferenceAbstractpeer-review


Aims: To determine the predictors of bereaved relative participation in a mortality followback survey. Method(s): A mortality followback survey 1516 bereaved relatives of cancer patients from four health regions in London (UK) were identified from death registrations covering one year period (2009/10). Sample stratified by death at home (n=368), hospital (513), hospice (512), nursing home (123). Respondents (n=596) and nonrespondents (920) were compared on socio-demographic variables using parametric (t-test) and non-parametric tests (Mann-Whitney, Chi2 and Fisher's exact tests), univariate and multivariate logistic regression. Result(s): Four factors - patient age (P< 0.001), place of death (P=0.021), informant gender (P< 0.001) and relationship to patient (P< 0.001) were found to be independently associated with participation. The odds of taking part in the study were highest if the patient was aged 90+ (OR 3.46, 95% CI 1.51 to 7.94) and the informant of death was female (OR 1.70, 95% CI 1.33 to 2.16). The odds of participating were lowest if the patient died in hospital (OR 0.62, 95% CI 0.46 to 0.84) and the informant of death was not a spouse/parent (ORs ranged 0.63 for siblings to 0.28 for others). Social deprivation was only marginally significant (P=0.053) and non-linear. Time following death (median 7 months), patient gender, country of birth, cancer type and health region were not independently associated with participation (P>0.05). Conclusion(s): Consistent with other studies, women and relatives of older patients were more likely to respond. In contrast to other studies, time following death and social deprivation did not influence participation. Relatives of people who died in hospital were less likely to respond. Further analysis of reasons for refusal may help identify ways to capture the views of less represented groups, either by increasing research participation in surveys or designing alternative studies.
Original languageEnglish
Number of pages1
Publication statusPublished - 7 Jun 2012


  • mortality
  • death
  • cancer patient
  • female
  • human
  • *palliative therapy
  • United Kingdom
  • registration
  • *cancer mortality
  • hospital
  • nursing home
  • logistic regression analysis
  • patient
  • social isolation
  • gender
  • neoplasm
  • Fisher exact test
  • health
  • hospice nursing
  • nonparametric test
  • sibling
  • Student t test


Dive into the research topics of 'Factors influencing participation in a cancer mortality followback survey: Palliative Medicine'. Together they form a unique fingerprint.

Cite this