Monitoring adverse social and medical events in public health trials: assessing predictors and interpretation against a proposed model of adverse event reporting.

Gwenllian Moody, Katy Addison, Rebecca Cannings-John, Julia Sanders, Carolyn Wallace, Michael R Robling

Research output: Contribution to journalArticlepeer-review

18 Downloads (Pure)


Background Although adverse event (AE) monitoring in trials focusses on medical events, social outcomes may be important in public or social care trials. We describe our approach to reporting and categorising medical and other AE reports, using a case study trial. We explore predictors of medical and social AEs, and develop a model for conceptualising safety monitoring. Methods The Building Blocks randomised controlled trial of specialist home visiting recruited 1618 first-time mothers aged 19 years or under at 18 English sites. Event reports collected during follow-up were independently reviewed and categorised as either Medical (standard Good Clinical Practice definition), or Social (trial-specific definition). A retrospectively developed system was created to classify AEs. Univariate analyses explored the association between baseline participant and study characteristics and the subsequent reporting of events. Factors significantly associated at this stage were progressed to binary logistic regressions to assess independent predictors. Results A classification system was derived for reported AEs that distinguished between Medical or Social AEs. One thousand, three hundred and fifteen event reports were obtained for mothers or their babies (1033 Medical, 257 Social). Allocation to the trial intervention arm was associated with increased likelihood of Medical rather than Social AE reporting. Poorer baseline psycho-social status predicted both Medical and Social events, and poorer psycho-social status better predicted Social rather than Medical events. Baseline predictors of Social AEs included being younger at recruitment (OR = 0.78 (CI = 0.67 to 0.90), p = 0.001), receiving benefits (OR = 1.60 (CI = 1.09 to 2.35), p = 0.016), and having a higher antisocial behaviour score (OR = 1.22 (CI = 1.09 to 1.36), p < 0.001). Baseline predictors of Medical AEs included having a limiting long-term illness (OR = 1.37 (CI = 1.01 to 1.88), p = 0.046), poorer mental health (OR = 1.03 (CI = 1.01 to 1.05), p = 0.004), and being in the intervention arm of the trial (OR = 1.34 (CI = 1.07 to 1.70), p = 0.012). Conclusions Continuity between baseline and subsequent adverse experiences was expected despite potentially beneficial intervention impact. We hypothesise that excess events reported for intervention-arm participants is likely attributable to surveillance bias. We interpreted our findings against a new model that explicates processes that may drive event occurrence, presentation and reporting. Focussing only upon Medical events may miss the well-being and social circumstances that are important for interpreting intervention safety and participant management.
Original languageEnglish
Number of pages16
Publication statusPublished - 30 Dec 2019


  • Safety Monitoring
  • Adverse events
  • Serious adverse event
  • Public Health
  • Clinical trials
  • Home visiting


Dive into the research topics of 'Monitoring adverse social and medical events in public health trials: assessing predictors and interpretation against a proposed model of adverse event reporting.'. Together they form a unique fingerprint.

Cite this