Determinants of presenteeism prevalence and propensity: Two sides of the same coin?

Abstract

Research on the determinants of sickness presence can be based on employees' number of sickness presence days or employees' propensity to "opt" for presenteeism in case of sickness. This study compares associations between potential determinants and presenteeism in cross-sectional data from a sample of Austrian employees by considering days of sickness absence and presence and an estimator for sickness-presence propensity as outcome variables. The sets of significant determinants for the number of sickness presence days and propensity were found to be overlapping, but not equivalent. Whereas determinants of the number of days can help to identify groups with high presenteeism prevalence, determinants of propensity can help to identify factors that restrict the decision process in case of sickness. With respect to possible interventions aimed at reducing presenteeism, results relating to both perspectives should be considered.

Sickness presenteeism: Are we sure about what we are studying? A research based on a literature review and an empirical illustration

“…These studies may be classified into two groups: (a) those based on the European Working Conditions Survey (EWCS), where the question used is “over the past 12 months did you work when you were sick?,” the answer being dichotomous (no/yes) . Some studies exclusively analyze this question, while others additionally take a second question into account, namely “if yes, how many working days?” The versions of the Korean Working Conditions Survey also use a dichotomous question similar to that of the EWCS, although with slightly different formulations depending on survey version; (b) those studies which use exactly the same question used by Aronsson, “has it happened over the previous 12 months that you have gone to work despite feeling that you really should have taken sick leave due to your state of health?,” as well as some other studies which use minor modifications of this question or other similar ones where the formulation explicitly investigates “how many times…” or “how many days…”…”

“…Nor is there a perfect consensus with respect to the population in which to estimate SP or, to put it another way, the denominator that should be used when estimating frequency measures. There are studies in which the population considered is all workers and others, noting that to be presenteeist the worker first must meet the condition of being “sick,” exclude “healthy” workers (considered “not at risk”) …”

“…We may consider SL as an adjustment variable, which in fact is done in most studies. Here however we are asking whether SL ought perhaps to play a role in defining the phenomenon being studied and, therefore, form part of the response variable, for example, the creation of a “presenteeism propensity” by dividing SP days by health events (ie, SL days + SP days) . Other studies define various combinations of SP and SL: for example, Gustaffson and Marklund classify workers in four groups based on their behavior in SP (zero to one episode vs two or more) and SL (<7 days, vs 1 week or more).…”

“…On the other hand, it would seem logical to exclude individuals who have been inactive or unemployed during a certain time, from SL and/or SP or due to the sum of various such aspects, since if during a certain time they did not work, nor were they at risk of presenting SP. Moreover, if the criterion for exclusion is based on SL, this would perhaps exclude workers with chronic pathologies, something which could be of interest, if the aim was that of studying the phenomenon of SP in terms of the exercising of rights.…”

“…The question used to record SP was “In the last 12 months, how many times have you worked even though you thought you should have taken sick‐leave due to your state of health?,” the answer is the total number of times. We operationally defined a worker as “not at risk” if during the preceding year they had no episode of either SP or SL (in other words they had not had any health problem sufficiently serious to justify not going to work), consistent with other studies . The covariates were sex, age, days of self‐reported SL, and self‐rated health (“In general, would you say your health is?”).…”