Prescriptions for Potentially Inappropriate Medication (PIM)

The term “potentially inadequate medication” (PIM) describes medications or active substances that should not be prescribed to older persons because of their adverse effects profile. Such relative contraindications should always be considered in the overall context and in view of comorbidities and further medications. The PRISCUS 2.0 list was published recently—a revised version of a PIM list for the German pharmaceutical market (1). When considering the number of older patients in the healthcare system for whom these medications may pose risks, this topic is more topical than ever. Thus far only few published data exist on the prevalence of PIMs and their development as well as on the PIM substances most commonly prescribed in primary care in Germany.

Methods

Our analyses are based on the SHIP Study (“Study of Health in Pomerania”)—a prospective population based cohort study that is being carried out in northeastern Germany (2). As PIM affects persons older than 65, younger people were excluded. We used data from the two cohorts SHIP-START and SHIP-TREND; for the baseline study we used data from the waves SHIP-START-2 and SHIP-TREND-0, which were conducted between 2008 and 2012. For the follow-ups we used data from the waves SHIP-START-3 and SHIP-TREND-1, collected between 2014 and 2016 and between 2016 and 2019. A total of n=874 subjects were included. The median time to follow-up was 6.2 years (interquartile range 5.0 years to 7.2 years). The following refers to this as the 6-year follow-up. All participants provided written consent. Both studies followed the recommendations of the Declaration of Helsinki and received ethics approval from the ethics committee of Greifswald University (approval No BB 39/08). Participants’ medications were recorded on the basis of their medication lists, and PIMs were identified on the basis of the PRISCUS list (3).

Results

We analyzed data from 874 study participants—451 (51.6%) were male and 423 (48.4%) were female. At the time of the baseline study, participants were aged between 65 and 90 years (median 70.7 years) and were prescribed between 0 and 16 active substances. The mean number of substances prescribed was 3.8 (standard deviation 2.8). PIM prescriptions were detected in a total of 97 patients (11.1%). 86 participants (9.8%) had been prescribed one PIM, nine participants had been prescribed two PIMs, and two participants (0.2%) had been prescribed three PIMs. At the 6-year follow-up, participants were aged between 69 and 93 years (median 76.8 years). At this time, participants were prescribed between zero and 21 actives substances (mean 4.8 (SD 3.2) active substances). At the time of the follow-up, prescriptions of PIMs were discovered in overall 99 participants (11.3%) (Figure), with 90 (10.3%) participants prescribed one PIM and nine participants (1.0%) prescribed two PIMs. At the time of the baseline study, 36 different active substances were prescribed and at the time of the 6-year follow-up, 33 different active substances were prescribed. The Table shows the 10 active substances that were prescribed most often.

Figure.

PIM prevalence and average number of prescribed medications at the time of the baseline study (2008 to 2012) and at the time of the follow-up (2014 to 2019)

Table. Overview of the most prescribed 10 active substances at the time of the baseline study and the 6-year follow-up.

Baseline study		Six-year follow-up
Active substance	n (%)	Active substance	n (%)
Solifenacin	10 (1.1 %)	Solifenacin	16 (1.8 %)
Flecainide	8 (0.9 %)	Doxazosin	15 (1.7 %)
Etoricoxib	8 (0.9 %)	Etoricoxib	10 (1.1 %)
Doxazosin	7 (0.8 %)	Flecainide	8 (0.9 %)
Nifedipine	6 (0.7 %)	Zolpidem	8 (0.9 %)
Zopiclone	6 (0.7 %)	Nifedipine	6 (0.7 %)
Diazepam	5 (0.6 %)	Trimipramine	4 (0.5 %)
Piracetam	5 (0.6 %)	Pentoxifylline	3 (0.3 %)
Pentoxifylline	4 (0.5 %)	Diazepam	3 (0.3 %)
Naftidrofuryl	4 (0.5 %)	Lorazepam	3 (0.3 %)

Discussion

In our study we did not detect any relevant reduction in the prevalence of PIM prescriptions after the introduction of the PRISCUS list. In our longitudinal study this was about 11%. Other studies have occasionally shown a fall in the prevalence of PIMs (1). The almost constantly high prevalence in our analyses can probably be explained with the fact (among others) that at both time points, the same participants were studied. At the time of the follow-up these were correspondingly older and therefore more multimorbid than at the time of the baseline data collection. Accordingly, the number of medications increased, as did the “risk” of being prescribed PIMs. Overall the prevalence we calculated is lower than that reported by Mann et al (1); the possible explanation is selection bias in the setting of our prospective cohort study.

Interestingly, solifenacin (a urologic spasmolytic substance) was prescribed most at both time points, and the number of prescriptions increased over time. A possible explanation is participants’ increasing age, but this affected none the less only 1.1% and 1.8% of the study population. Furthermore, the sedatives zopiclone, diazepam, zolpidem, and lorazepam were among the 10 most commonly prescribed PIMs; other studies also showed high PIM prevalence rates for sedatives (4). Overall, these prescriptions were rare in our study population; because of the sometimes serious adverse effects, such as increased risk of falls, alternative medication—with, for example, pipamperone or mirtazapine—should still be considered. Similarly, etoricoxib was prescribed in several cases, although the indication should be critically reviewed because of the increased risk of gastrointestinal bleeds, and the drug should possibly be substituted with paracetamol or transdermal analgesic patches. The antihypertensive drugs doxazosin and nifedipine were also among the most commonly prescribed PIMs. In poorly controlled arterial hypertension, they may be indispensable in case of doubt, but wherever possible, different antihypertensive drugs should be used in preference.

In sum, prescriptions of PIMs from highly diverse groups of active substances were identified in about one in every 10 persons >65, which poses a challenge for doctors in almost all specialties. Such prescriptions are not always avoidable, but handling them sensibly can contribute to patients’ safety.