Quality of drug treatment in older people from a medical primary care perspective and according to indicators

Abstract: Background: Drug treatment in older people is particularly complex; the burden of disease increases by age, and patients in this age group are more sensitive to drug effects. As a knowledge resource for clinicians as well as for benchmarking and pharmacoepidemiological research, indicators of prescribing quality for older people have been established, including potentially inappropriate medications (PIMs) and potential prescribing omissions (PPOs). Furthermore, an indicator linked to the coding of medication reviews in patients ≥75 years has been used for monetary compensation. The overall aim of this thesis was to investigate the quality of drug treatment in older patients from a medical primary care perspective, and according to indicators of prescribing quality. Methods and results: In all, 625 patients were included, encompassing a pilot study (I), a cross-sectional study (II), a validation study (III), and a reliability study (IV). All patients were ≥65 years of age; 425 represented consecutive patients in primary health care and 200 were hip fracture patients included in a previous interventional study. In study I, we found that two thirds of the patients had appropriate drug treatment from a medical perspective, and about half of all patients had at least one medication review recorded over the last year. In study II, more patients without a recorded medication review, compared with those with such a code registered, had adequate drug treatment; no further action related to the medication was medically justified before the next regular consultation as determined by two blinded specialist physicians in consensus (73% vs 63%, P=0.047). No difference was found for patients ≥75 years of age, the age limit for remuneration, and a multivariate logistic regression revealed that this factor was conspicuously associated with a medication review being recorded. In study III, we showed that the majority of patients (86%) appeared to have ≥1 PIM or PPO detected by three sets of indicators of prescribing quality, but only one in seven PIMs/PPOs was clinically relevant for the individual patient, and half of these were not prioritised for medical action before the next visit. For instance, the most frequent PIMs included proton pump inhibitors and drugs for insomnia, prioritised for medical action before the next visit in one out of 76 cases and three out of 58 cases, respectively. Sixty-three percent of patients with ≥1 PIM/PPO had adequate drug treatment from a medical perspective. In study IV, we found a weak inter-rater agreement regarding the detection of PIMs/PPOs according to an established set of PIMs/PPOs, within two physician pairs in two patient cohorts. Further analyses revealed that methodological issues, including the use of medical records instead of prepared cases and ambiguities in the criteria wordings, may have contributed to these findings. Conclusions: Applying a medical primary care perspective, drug treatment in older people is in general adequate, but the indicators of prescribing quality studied in this thesis appear not to reflect adequacy in a reliable and valid way. Further, the recording of a medication review seems not to be associated with adequate drug treatment but rather to the age limit for remuneration in primary health care. In conjunction, these results suggest caution in interpretation when these indicators are used to measure quality of drug treatment in older people. Our findings, including information regarding drug treatment in older people that may deserve more attention, and other information that does not, could be useful in clinical practice, in educational efforts, and in research and future development of indicators of prescribing quality.