Use of PIM decreased in all the variables, age, gender, number of chronic conditions and polypharmacy during a national information campaign to reduce PIM. However, the decrease was more evident in women, patients with polypharmacy and patients with two to four chronic conditions. The group that had the highest probability to deprescribe PIM during the study period was patients with polypharmacy and high number of chronic conditions.
The positive trend of the reduced prevalence of PIM users found in this study corresponds with results from other reports in Sweden during the same time period [28-30].
In 2005 the prevalence of PIM was found to be 17% in a Swedish older adult population and a national comparison showed that use of PIM had decreased by 44% between 2005 to 2014 [31, 32]. Use of PIM and polypharmacy is associated with increased risk for ADEs and hospitalisation [9, 33].
In our study, the prevalence of polypharmacy stayed relatively stable, but number of chronic conditions increased. The fact that polypharmacy did not increase significantly while the number of chronic conditions increased is an interesting finding. One could think that if multi-morbidity is increasing that polypharmacy would follow. However, the use of medications did increase, just not polypharmacy in comparison with the rest of the population.
The decrease in PIM in this study was not paralleled by a decrease in polypharmacy. The most common drug classes in patients 75 years and older with polypharmacy are not PIM (according to our definition) but cardiovascular drugs (including antithrombotic agents), analgesics and psychotropic drugs [34]. These are also the most commonly used drugs in adverse drug events, such as bleeding or bruising, which are associated with antithrombotic agents, or dizziness and unsteadiness due to psychotropic medicines [4].
It can be stated that based on this single quality indicator, the use of PIM has improved and thereby the quality of medication treatment in older adults. However, it does not affect the total quality of medications use. The information campaign was a success as regards that it reduced the use of PIM, especially in patients with high number of chronic conditions and polypharmacy. However, it did not reduce polypharmacy, which is also an important factor for quality in medication use in older adults [20, 35-37].
The results in this study show that the use of a clear and simple quality indicator as decreased use of PIM can improve the quality of medication treatment in older adults. However, to affect other factors of importance for the quality of medication treatment, a combination of quality indicators may be better to use. For example, the STOPP criteria, a collection of quality indicators, reduced the number of ADEs when implemented in a hospital setting in a study from Cork University Hospital [38]. The complete collection of quality criteria in “Quality indicators for good drug therapy in elderly” from the Swedish National Board of Health and Welfare can be used in the same way [13]. The effect is more complex to evaluate on a population level, but the clinical effect in the individual is greater.
Strengths and Limitations
The definition of PIM used in this study is stricter in its definition and includes fewer drugs and drug classes than other definitions [13, 15]. For example, we do not include nonsteroidal anti-inflammatory drugs (NSAID) or cardiovascular drugs except for disopyramide. However, our definition from the Swedish National Board of health and Welfare is commonly used in Sweden as an indicator for quality of drug treatment in older adults, both nationally and by county councils and is therefore relevant in this setting. On the other hand, this means that our results cannot be directly translated to other settings where the definition of PIM is broader.
The county council’s register, used in this study to identify use of medication, includes medications that are prescribed, and pharmacy dispensed for all inhabitants in Blekinge. Use of illegal drugs or over the counter drugs are not included in this study. In 2011, the Medical Products Agency in Sweden presented a study that indicates that 11% of the Swedish population had at some time bought prescription drugs from non-approved pharmacies [39]. The method to construct a medicine list on dispensed prescribed drugs from the inclusion date for the cohorts and three months back, allowed us to determine, as closely as possible to index date, as to what the patient was using. The limitation of this method is the possibility of underestimating the use of medications prescribed to be used as needed. This because they are dispensed more rarely than every three months. The patient’s compliance, i.e. if they were using their medications as prescribes, were we not able to take into consideration when determining use of PIM. However, the method used, to identify use of medication in this study, is validated and the time period of three months has been found to be the most optimal [25]. Factoring in the Swedish system that has a high cost threshold, there is a limited risk of hoarding medications [23].
Number of chronic conditions, that are used in this study to measure multimorbidity, is dependent on the quality of registration of diagnoses in the medical records [27]. The recording of diagnoses in this study has not been validated. However, we used registered diagnoses from a two-year period from both primary- and secondary care to get as close to total coverage as possible. There are other multimorbidity estimates that are constructed by giving different diagnoses a weight as to how much the diagnosis contributed to need of care or cost [41]. In our definition, all chronic conditions contribute equally to the estimate, i.e. an expression of the complexity of a patient’s morbidity and their need of care.
We used two different cohorts in this study instead of following one cohort over time when analysing prevalence of PIM and polypharmacy in relation to studied variables. If we had used only the individuals present in both cohorts (78%) the results would have been affected by the fact the population had aged.
Blekinge County is a small county in Sweden, both in terms of population and area, and has a relatively simple organisation of health care service, which makes it easy to include data from primary care centres, both public and private, and from secondary care. Our results are applicable to populations with older adults in similar settings.