In partnership with ScienceOpen

600 word lay summary created by Science Impact – Stopping smoking studies: Why people quit

Withdrawal of participants from stopping smoking studies is a common phenomenon and is problematic in that it can have serious methodological and clinical consequences for the study in question. For example, participants dropping out can prevent a statistical conclusion being reached, and damage a study’s validity. Ultimately, this limits the clinical applicability of potentially useful interventions.

One option to overcome this is for researchers to recruit and enrol more people in a study, but this means increased cost and manpower. In order to discover more feasible potential solutions to the issue, a team of researchers set out to explore rates of attrition in stopping smoking studies and investigate reasons behind the phenomenon. Their hope is that this could inform strategies to address the problem in the future and prevent its consequences.

The team performed a descriptive systematic review of existing studies evaluating the effectiveness of stopping smoking interventions. They looked at both pre-inclusion and post-inclusion attrition, as attrition poses the same methodological problems for researchers irrespective of when it takes place.

The researchers used several databases to search for relevant studies. In order to be included, a study had to have been published between 1980 and 2015; be an experimental or quasi-experimental study; be pharmacological, educational or behavioural; have targeted adult smokers; have reported the rate of attrition, or allowed for it to be calculated; and, finally, have explored factors associated with attrition.

Of the 189 articles that met these criteria, only 10 examined factors influencing attrition in smoking cessation intervention studies. One of these studies was excluded as the attrition rate was not explicitly reported and there were no data available to calculate it. Data were extracted on study characteristics, intervention characteristics, attrition rate, and factors influencing attrition and, in order to facilitate data abstraction and synthesis, this information was incorporated into a table.

The researchers found that attrition rates ranged between 10.8 per cent and 77 per cent, and that a small number of demographic, clinical, behavioural, health, health-related beliefs, and logistical factors were related to attrition. According to the team’s findings, participants’ age, gender, education, and nicotine dependence are related to attrition in smoking cessation intervention studies. For example, the researchers found that younger age was consistently associated with attrition, but that what exactly leads younger smokers to withdraw is not well understood. Three studies found that men are likely to withdraw from smoking cessation intervention studies, whereas one study reported that women are more likely to drop out.

Smokers with a low level of education were found to be likely to withdraw, and smokers with high levels of nicotine dependence were consistently found to drop out, even though this is the demographic most in need of stopping smoking interventions. Additionally, participants who received frequent email reminders were found to be more likely to drop out.

The researchers believe that exploration of additional logistical factors such as transportation and childcare issues is warranted. They also argue that methods of treatment assignment that account for participants’ preferences could reduce attrition.

The researchers identified the following strategies to minimise attrition throughout various phases of a study: modifying interventions to fit participants’ literacy levels; tailoring cessation interventions by assessing participants’ level of nicotine dependence and providing quitting smoking strategies that correspond with that level; ensuring flexibility in treatment and research activities; and provision of childcare. They conclude that the effectiveness of these strategies should be investigated in future research.