1/ Why measuring polypharmacy and on what method is based your study?

The increase in life expectancy is associated with an increase in years of life, and often with several chronic diseases causing an accumulation of treatments. From an individual perspective, it has been shown that polypharmacy exposed to an increased risk of falls, hospitalization and death, and, moreover, causes a burden on the patient which can affect his quality of life. From the perspective of the health system, the issue is also economic: with the additional costs generated by the consumption of unnecessary drugs comes the iatrogenic cost in terms of hospitalisations, physician visits and additional medication. In this context, public policies aiming at reducing iatrogenic risk in the elderly have been introduced in recent years. Their assessment requires easy-to-produce indicators.
We have studied two indicators: the "cumulative" polypharmacy indicator which takes into account all drugs administered over a given period, and the "continuous" polypharmacy indicator which focuses on drugs taken regularly over a prolonged period of time. The innovative aspect of our work is the consideration of combination drugs and quarterly packs in the calculations. Indeed, if the goal is to reduce the risk of adverse events, we'd seek to reduce the number of molecules to which the patient is exposed, whether these molecules are associated or not in a tablet. Moreover, as treatments in question are often related to chronic diseases, it is therefore important to take into account the specific packs in order to space out drug dispensing.

2/ What are the main results of your study?

The prevalence of polypharmacy at the 10-drug threshold in France in individuals age 75 and over for the continuous and cumulative polypharmacy indicator is 33 % and 40.5 %. The prevalence of polypharmacy increases by 7 points when considering the total treatments during the year and not only treatments taken chronically. So it seems that the majority of persons considered polymedicated at the 10-drug threshold are polymedicated because of chronic disease treatments. Indeed, amongst 40 % of beneficiaries classified as polymedicated with the cumulative indicator, 77, 5 % are considered polymedicated with the continuous indicator.
Another important result of our study, the consideration of combination drugs modifies the prevalence of polypharmacy: whether using the cumulative or continuous indicators, the consideration of combination drugs (and quarterly packs for the continous indicator) increases by 6 points the prevalence of polypharmacy.
Amongst the 15 most frequent therapeutic (concerning most beneficiaries), 11 are common to the cumulative and continuous indicators. These therapeutic classes are most often used in the treatment of chronic diseases and rarely in the treatment of acute conditions.
Finally, associated factors (demographic, geographic, individuals covered by the long-term illness scheme…) do not differ according to the indicators and do not vary with the consideration of combinations and quarterly packs.

3/ What lessons can be learned from your study in terms of public policy and what will be the next developments on this issue?

The implementation of actions aimed at reducing polypharmacy is limited to acute treatments or treatments of acute diseases. The continuous indicator, which targets chronic disease treatments, seems therefore most suitable for monitoring public policies.
This work allowed to assess the bias associated to the non-consideration of combination drugs, notably in the prevalence calculations of polypharmacy. Taking into account combinations require significant work, which is not easily feasible. In this case, having a quantitative knowledge of the bias generated could be an interesting element. In order to allow other users of the National Health insurance databases to easily calculate these indicators taking into account combinations, macros were shared on the portal users of the National Inter-regimes Health Insurance Information System (SNIIRAM).
A polypharmacy indicator remains however only a number and cannot be used in the identification of risky drug use. Pairing a polypharmacy indicator with algorithms identifying associations of risky drugs could be considered as a global monitoring system of drug use in the elderly.

Interview by Anne Evans

All interviews...