With the global population aging, the total number of people aged 60 years and older in the world is expected to reach 2 billion by 2050. China is the most populous country in the world, and the older population is also the largest (Jia et al., 2020). Older adults are more likely to suffer from multiple diseases, especially chronic diseases requiring complex treatments, such as taking many different medicines (Cojutti et al., 2016). Polypharmacy (defined as more than five medicines) is associated with the prescription of inappropriate medications, and a growing body of evidence links polypharmacy with negative outcomes (Field et al., 2001; Ferner and Aronson, 2006; Maddison et al., 2011; Weng et al., 2013; LeBlanc et al., 2015).

However, alterations in age-related pharmacokinetics and pharmacodynamics of older adults have led to an increased risk of drug–drug interactions and drug–disease interactions (Fried et al., 2014; Payne, 2016). Cancer patients are particularly prone to unintended consequences of polypharmacy because chemotherapy may carry a risk of drug–drug interactions and adverse drug events, which may include chemotherapy-related toxicity (Maggiore et al., 2014; Woopen et al., 2016). Some studies have shown that older cancer patients could suffer from a higher rate of comorbidity, frailty, and geriatric syndrome, putting them at high risk of polypharmacy and inappropriate medication use (Wildiers et al., 2014; Koczwara et al., 2022).

Potentially inappropriate medication (PIM) is a public health issue that can be defined as medications that should be avoided and may outweigh the expected clinical benefit, such as adverse drug events, hospitalization, disability, and economic burden (Hyttinen et al., 2016; Muhlack et al., 2017; Wallace et al., 2017). The American Geriatrics Society (AGS)/Beers criteria were the first expert consensus on geriatric PIM (Beers et al., 1991). The AGS, through an expert US-based panel, has undertaken the task of regular review and updating of AGS/Beers criteria, which are now in their sixth iteration (American Geriatrics Society Beers Criteria® Update Expert Panel, 2019). There were some substantial changes in the categories, and some medications were dropped or added. Because Beers criteria were not organized according to physiological systems, University College Cork organized experts from many disciplines to formulate the screening tool of old persons’ prescriptions to alert to the right treatment (STOPP/START criteria) through the Delphi method, and the second edition was updated in 2014 (O'Mahony et al., 2015; O'Mahony, 2020). Two criteria have been widely used in PIM use application surveys in communities, clinics, and hospitals worldwide. China formulated the criteria for judging the potentially inappropriate medication use of older adults by an expert panel in 2017, including medication risk and medication risk under disease status (Rational Drug Use Branch of Chinese Association of Geriatric, 2018). These country-specific criteria were divided into high-risk and low-risk medications according to experts’ evaluation.

Some previous reports examined PIM use in older Chinese patients based on the three criteria. However, no study has specifically reported on the concordance among the three criteria. The prevalence and the risk factors associated with PIM use according to the three criteria in older Chinese cancer patients are unclear. The concordance of different criteria often led to large differences in the results. Besides, country-specific and non-country-specific criteria significantly impact PIMs in older cancer patients. Therefore, in this study, we extracted prescriptions of cancer outpatients treated at tertiary hospitals in Chengdu, China. PIMs were screened based on the 2017 Chinese, 2019 AGS/Beers, and 2014 STOPP criteria. The concordance among the three PIM criteria was calculated, and the prevalence and the risk factors associated with PIMs were explored. It is hoped that this study will provide relevant evidence for follow-up research.

To the best of our knowledge, this is the first study assessing the concordance of three PIM-detecting tools—the 2017 Chinese criteria, the 2019 AGS/Beers criteria, and the 2014 STOPP criteria—in older Chinese cancer outpatients. Although these criteria were developed for different populations and with different aims, they are the most commonly used in older Chinese patients. Because multiple comorbidities are frequent among older cancer patients, a tool focusing on cancer outpatients should be implemented to alert doctors to an eventual PIM prescription. Our study found that the 2017 Chinese and the 2014 STOPP criteria indicated poor coherence, whereas the 2019 AGS/Beers criteria showed moderate concordance with the other two criteria, which was a little different from another study on Chinese older inpatients (Ma et al., 2018). Moreover, a Portuguese study performed in inpatients 65 or more years of age showed poor concordance among the 2019 AGS/Beers criteria, 2014 STOPP criteria, and the EU(7)-PIM list (Perpétuo., 2021). The low concordance between different criteria highlights the need to develop special PIM-detecting criteria for older cancer patients exposed to many PIMs and reinforces the fact that older cancer outpatients are also at risk of PIM. This will provide a basis for rational drug use for cancer patients and reduce outpatient prescription expenditure. The poor concordance between the Chinese and the STOPP criteria can be due to the applicability requirements of each list. The overlap between the Beers criteria and the other two criteria regarding medication risk irrespective of conditions was relatively high. However, the Chinese criteria contained clopidogrel and mixed insulin not included in the Beers criteria. In order to determine one PIM with the STOPP criteria, it is imperative to know the entire medication history and clinical information of the patient. These reasons may lead to moderate concordance between the Beers criteria and the other two criteria.

China is currently the country with the largest older cancer population in the world, and cancer as a chronic disease places a heavy burden on the elderly. Older cancer patients can suffer from a higher rate of comorbidity, frailty, and geriatric syndrome, putting them at a high risk of polypharmacy and PIM use (Pamoukdjian et al., 2020; Kleckner et al., 2022). To the best of our knowledge, this is the first cross-sectional study on the prevalence and risk factors for PIM use in Chinese older cancer outpatients according to the three criteria. The prevalence of PIM use was 34.37%, 32.65%, and 15.96%, according to the 2017 Chinese, 2019 AGS/Beers, and 2014 STOPP criteria, respectively. There is little difference between the 2017 Chinese and 2019 AGS/Beers criteria. However, the prevalence of PIM use of the 2014 STOPP criteria was lower than the other two criteria. According to the 2017 Chinese criteria, to consider the medicine as a PIM, it is only necessary to know the status of medication and disease in older patients. In addition, Chinese criteria were made based on drug utilization of the older Chinese population, so it is more suitable for Chinese individuals. The AGS/Beers criteria judge each medicine as a PIM based not only on the medication profile of a patient but also on the pathologies of the patients, as well as the laboratory results (O'Mahony et al., 2015). In order to apply the STOPP criteria, it is imperative to know the entire medication history, clinical information of the patient, and laboratory (O'Mahony et al., 2015; O'Mahony, 2020; Perpétuo et al., 2021). Based on the 2019 AGS/Beers criteria, our study found that the prevalence of PIM use among older Chinese cancer patients was 32.65%, which was lower than the prevalence of 80.4% reported by a study on Korean cancer patients according to the 2019 AGS/Beers criteria (Suh et al., 2021). The older Korean patients received anti-neoplastic therapy with emergency department (ED) visits, the prevalence of polypharmacy in the patients was observed in 80.4%, and the prevalence was 22.53% in our study. Taking more medications was the reason for the higher prevalence of PIM use compared to our study. Based on the 2014 STOPP criteria, our study found that the prevalence of PIM use among older cancer outpatients was 15.96%, which was lower than Japanese with a prevalence of 31.9% (Hakozaki et al., 2021). Older advanced non–small cell lung cancer (NSCLC) patients and those on oral molecular-targeted anticancer agents were included in the study. According to our research, the prevalence of PIM use in lung cancer patients was higher. The high prevalence of PIM use is that older cancer outpatients are usually in serious condition both physically and mentally, and the willingness of patients to take medicine is relatively strong, not only for antitumor drugs but also for analgesic drugs and sedative-hypnotic drugs. Another potential reason was that the adverse outcomes in older cancer patients were highly associated with PIM use, and the poor clinical outcome of cancer patients will further aggravate the prevalence of PIM use (Mohamed et al., 2020; Chen et al., 2021).

In our research, the most frequent PIM in Chinese older cancer outpatients was estazolam, according to the three criteria. Sleep disorder is common with advancing age and affects 36%–70% of older adults (Hishikawa et al., 2017; Patel et al., 2018), and it is further aggravated in older cancer patients. Consequently, estazolam is a benzodiazepine frequently used by older Chinese cancer patients. However, benzodiazepines are also linked to risks of mortality, falls, fractures, and depression among older adults (Kripke et al., 2002; Stone et al., 2008; Yaffe et al., 2014). Therefore, the risk of this category of medication use should be further evaluated for older cancer patients.

According to the results of logistic regression analysis, PIM-associated factors were the same among the three sets of criteria; older cancer outpatients who were ≥80 years of age, had more diseases, had polypharmacy, and had an irrational use of drugs and those who had lung cancer were more likely to receive PIMs. Furthermore, compared with other identified factors, polypharmacy is the most strongly associated independent risk factor. Patients with polypharmacy had more than two to three times the risk of PIM use compared with patients with one to four medications. In this study, the polypharmacy of older cancer outpatients was 22.53%, which is slightly little lower than the result of our other study (Tian et al., 2021), and this was similar to the results of Hsu et al.'s study, in which polypharmacy prevalence was lower in those with than without a cancer history (Hsu et al., 2021). Older cancer patients with age more than 80 generally have worse health and more multimorbidity than the general cancer population of older adults, and they are more likely to be exposed to PIM use (Lai et al., 2018). Our study found that, with the increase in multimorbidity in Chinese older cancer patients, the risk of PIM use gradually increased. This phenomenon is similar to older Chinese patients with other chronic diseases in some studies (Li et al., 2021; Zhao et al., 2021). The growth was more obvious with the 2014 STOPP criteria, as the PIM use of these criteria was more affected by the disease. In addition, unreasonable prescribing carries a higher risk of PIM use in Chinese older cancer outpatients. However, with the increase in prescription expenditures for cancer patients, the prevalence of PIM gradually declined. This was because the high expenditure on cancer prescriptions was mostly due to the use of antitumor drugs. However, the three criteria rarely involve antitumor drugs. Among all cancer diseases, only lung cancer was associated with PIM use. One study showed that at least half of patients with lung cancer have comorbidities, which would increase the risk of PIM use (Pluchart et al., 2021). Through these results, we suggested reducing unnecessary medications and performing medication reconciliation carefully for older cancer outpatients with taking multiple medications from the doctor or the pharmacist. At the same time, the criteria could be more refined according to the risk factors, such as the formation of special criteria for the outpatients who were ≥80 years of age and older lung cancer patients. This will further improve the feasibility and accuracy of the criteria.

Several limitations should be noted in this study. It was an observational study conducted in China, which is likely to cause some deviations in the results. These results need to be further confirmed by multicenter clinical trials. Second, there are no follow-up data for these older cancer patients when investigating PIM use by electronic medical data, so the correlation between PIM use and further clinical outcomes is not known. Finally, the patients attending outpatients of tertiary hospitals were the main focus of the study, and cancer outpatients who were in nursing homes and communities were not evaluated.

This study investigated the use of PIMs in older cancer outpatients with multimorbidity in Chengdu based on the 2017 Chinese, 2019 AGS/Beers, and 2014 STOPP criteria. The results showed that the prevalence of PIM use was high in Chinese older cancer outpatients; poor-to-moderate concordance among the three criteria was observed; and age ≥80, more diseases, polypharmacy, irrational use of drugs, and lung cancer were risk factors for PIM use.