Introduction
Older patients frequently suffer from multiple comorbidities, increasing the risk of polypharmacy and therefore drug-related problems (DRPs) [
1]. According to pharmaceutical care network of Europe (PCNE),
DRP is defined as, “an event or circumstance involving drug therapy that actually or potentially interferes with desired health outcomes” [
2]. Fall-related injuries are a major health issue in older people, as they are not only associated with additional rehabilitation, medical, and social complications, but also with a significant economic burden on the health care systems [
3]. When geriatric patients with fall-associated injuries are admitted for surgical treatment, their risk for complications is increased due to underlying comorbidities, polypharmacy, frequent treatment changes during hospitalization, and limited expertise on complex pharmacotherapy among most surgeons. As such, geriatric counseling is recommended for older patients on an orthopedic ward [
4,
5].
Several studies have demonstrated that a multidisciplinary approach can improve outcomes in terms of hospital readmissions, and quality of life (QOL), of these patients [
2,
6]. An orthogeriatric co-management (OG-CM) model is a sophisticated model for the management of frail patients in which a geriatrician is integrated into the orthopedic ward, to manage the patient together with the orthopedic surgeon from admission to discharge [
7]. This model demonstrates an increase in quality of care, as evidenced by the increased number of diagnoses of comorbidities resulting in having less readmissions, which is beneficial for health care systems [
7]. Besides the geriatrician, the clinical pharmacist (CP) can help to optimize pharmacotherapy [
2], by means of medication reconciliation (MR), review, identifying fall-risk increasing drugs (FRIDs), counseling of the patient or caregiver and post-discharge follow-up [
8,
9]. Transmural communication to primary care providers (PCPs) such as general practitioners and community pharmacists is essential, especially at crucial moments, such as care transitions [
10]. There is a growing body of literature that recognizes the importance of interdisciplinary collaboration to implement drug-specific recommendations. The general practitioner (GP), for example, plays a key role as he maintains an overview of all the patient’s prescribed medication and comorbidities. However, only a few studies have included the GP at discharge to discuss hospital-based recommendations [
2].
Therefore, the aim of this study was to investigate the impact of pharmaceutical interventions (PIs) with regard to the DRP reduction rate after discharge, in geriatric patients admitted to the orthopedic ward.
Methods
A prospective, monocentric interventional study with a pre-post design was conducted on the orthopedic ward (29 beds) of the University Hospital of Brussels, a 721-bed tertiary hospital in Belgium. Patients in the pre-group received usual care including the OG-CM model (October 1st to December 31st 2021), while patients in the post-group received the multidisciplinary approach including pharmacist-led interventions (January 1st to March 31st 2022). A pre-post approach was chosen to prevent contamination bias in the usual care group.
Study population
Inclusion criteria were patients aged 65 or more, patients with an orthopedic problem admitted to orthopedic or other surgical wards, either through the emergency department or after ambulatory specialist referral, hospitalization for > 48 h, and Dutch or French speaking. Exclusion criteria were logistical reasons (e.g., isolation due to COVID), refusal of informed consent and a setting of palliative care.
Study procedure
During the control period, usual care concerning drug therapy of the included patients was documented; no PIs were carried out. The ward’s residents were responsible for MR, review and pharmaceutical follow-up at discharge. There was a full-time geriatrician collaborating with surgeons (OG-CM model) who focused on older patients as part of an integrated co-management strategy (orthopedic and trauma surgeons, a member of the Geriatric Liaison Service, a dietician, a physical therapist and a social nurse). According local policy, drug-related recommendations were preferably made upon discharge due to the patient’s short length of stay (LOS). Treatment changes only occurred if deemed necessary by the geriatrician.
In the intervention period, a CP was added to the OG-CM model. The pharmacist-led intervention included MR (conducted with the help of PCPs such as the community pharmacist (CoP) and GP), medication review during hospitalization, optimizing patient counseling at discharge and post-discharge follow-up of patients. At discharge, the identified DRPs were discussed with both the geriatrician and surgical resident to determine which interventions could be proposed to the GP in the discharge letter. The CP provided a transitional pharmaceutical care plan which was integrated in the discharge letter, as well as patient counseling [
11]. At the end of the hospitalization, the CP reviewed the medication for remaining (potentially avoidable) DRPs. This was done using explicit medication assessment tools (GheOP
3S tool (Ghent Older People’s Prescriptions community Pharmacy Screening) [
12] and Stockley’s Interactions Checker [
13]). The CP contacted the GP at discharge to discuss changes and he proposed a follow-up plan for drug-related interventions.
One month after discharge, the patient’s GP was contacted again to reevaluate the home medication and to determine unresolved DRPs. If the GP could not be reached, the current medication was obtained by contacting the CoP.
Patient characteristics (age, gender, type of residence before admission, reason for admission, Charlson Comorbidity Index (CCI) [
14], medication fall-risk score [
15]) and data concerning the hospitalization were collected from the electronic patient records in the hospital’s information system (PrimUZ®). Medication-related information (MR, data concerning patient counseling and follow-up) was also documented during both periods. All data were registered in an online database on the REDCap® platform (Vanderbilt University, Nashville, TN, USA).
Outcomes
The DRP reduction rate was calculated and compared in the control- and intervention-group as the primary outcome, by determining the difference between the amount of DRPs at discharge and the amount one month later.
Additionally, GP acceptance rates were determined for the proposed PIs in the intervention-group, differentiating between interventions accepted immediately at discharge and those accepted after one month. Readmissions within one month following discharge were documented. An estimation of the clinical impact (CI) of the proposed PIs, was done using the methodology as proposed in the Clinical, Economic and Organizational (CLEO) tool [
16]. All the PIs were scored by an independent geriatrician and CP. Clinically significant PIs were those with a CI ≥ 2 C [
16].
Data analysis
Data are presented as means and standard deviation (SD) or median with interquartile range (IQR) as appropriate. Frequencies (percentage) were calculated for categorical variables. The unpaired
t-test and Mann-Whitney U-test were used to compare continuous variables and the Chi-square test for frequencies with Bonferroni correction where needed. The impact of the intervention on DRP reduction rate, as well as other patient- or drug-related characteristics was explored in a stepwise multiple linear regression analysis. The final model was tested for multicollinearity, homoscedasticity and normality of residuals.
P-values less than 0.05 (two- sided) were considered statistically significant. A weighted Cohen’s Kappa (κ
w) analysis was done to determine the inter-rater agreement between the two healthcare providers scoring the CI of PIs [
17]. All data were analyzed using IBM SPSS Statistics® version 28.0 (IBM Corporation, Armonk, NY, USA).
Discussion
The aim of this study was to investigate the impact of PIs, with regard to reducing DRPs in orthogeriatric patients after discharge. This was done by transmural communication of the proposed PIs at discharge to GPs in order to increase the acceptance rate of the proposed PIs. A multidisciplinary approach was used consisting of the incorporation of a CP in the already existing OG-CM model at the orthopedic ward [
7]. The collaboration between the geriatrician and orthopedic surgeons, has proven to improve the quality of care for orthogeriatric patients.
The most common DRPs identified in our study were similar to those in a general geriatric ward [
12]. Orthogeriatric patients have a high need for pain relief, which resulted in a number of DRPs involving drugs causing constipation, combinations of anticholinergic drugs, and omission of laxatives in concomitant opioid use. Compared with the study by Kympers et al., the use of PIMs, such as FRIDs, was more frequent. This could be explained by the fact that this study targeted geriatric patients with fall-related injuries.
The multidisciplinary approach in the intervention period was considered successful as it significantly reduced the number of DRPs compared to the control period. Approximately 44% of the existed DRPs were resolved after the PI. This number is lower compared to other studies (58.9–68.3%) [
18]. Variations may be explained by differences in the included population, i.e., no age restrictions, with older patients at higher risk for DRP development due to various comorbidities and age-related changes in pharmacokinetics and pharmacodynamics [
2]. Additionally, the instruments used to identify DRPs can have an impact. Kympers et al. also used the GheOP
3s tool to identify DRPs and reported a similar number of DRPs per patient (median = 4) [
12], compared to a median of 3 in our study. These numbers are consistent with reported numbers in literature (1.3–3.3 DRPs per patient) [
2,
19].
A statistically, significant decrease of 1 DRP per patient was observed in our study in the intervention period compared to the control period. The clinical significance cannot be precisely established but we can make an estimate by looking at which DRPs were resolved one month after discharge in the intervention period (Table
2) and the assessment of the CI of these PIs by the independent geriatrician and CP. For example, we mainly saw a reduction in DRPs such as combination of QT-prolonging drugs, intake of antidepressants > 1 year, intake of PPIs > 8 weeks and combination of drugs leading to increased bleeding risk. All these PIs, for which PIs was proposed, were considered as having a moderate of major CI by the independent geriatrician and CP.
Besides the intervention itself, the number of DRPs at discharge appeared to significantly increase the DRP reduction rate. In other studies, factors such as number of prescribed drugs on admission, CCI and LOS, were shown to be associated with an increase in DRPs [
20,
21]. In this study, no association was observed between the increase in DRPs and these determinants. This may be due to differences in the study population, as the other studies mainly focused on patients admitted to a geriatric internal medicine ward. Baseline patient characteristics of both groups in our study were similar, except for the patients’ LOS. About half of the included patients in the intervention group were discharged to a rehabilitation facility, while in the control group only 1 in 3. This could possibly be explained by the longer waiting times for rehabilitation centers due to the COVID pandemic in the intervention period.
In contrast, recent hospitalization and number of drugs on admission, did not result in a decreased DRP rate at one month post-discharge. A possible explanation might be that recently admitted patients were at lower risk of having multiple DRPs, as these DRPs could have been resolved during their previous hospitalization. Remarkably, a higher fall-risk score (≥ 6 higher risk for fall), might be associated with a decrease in DRP reduction rate (Table
3). The timely identification and deprescribing of FRIDs should be of utmost importance in this setting as part of a multifactorial fall-prevention strategy [
9]. We hypothesize that GPs may be hesitant to accept interventions regarding these drugs because optimizing drug therapy through deprescribing is very intensive and time-consuming. Motivation of PCPs, knowledge, lack of time for deprescribing, and miscommunication between specialists, PCPs and patients can be either facilitators or barriers of deprescribing [
22]. Optimizing the deprescribing process, especially for drugs in which the potential harms outweigh the potential benefits, may improve outcomes for these patients. PCPs involvement is crucial in order to achieve a sustainable DRP reduction [
11]. Other professionals such as the CoP can also be included to support the GP with DRP follow-up, thus contributing to improvement of pharmaceutical care. A study with a multifaceted approach and patient-centered and primary care directed intervention, proved to be effective for deprescribing as more than 90% of the older patients agreed to discontinue unnecessary drugs when recommended by their GP [
23]. Further research is needed, focusing on these drugs to develop specific interventions that combine explicit and implicit approaches, with patient-centered decision making [
24].
Furthermore, the physician acceptance rate of the proposed interventions was high (71%), though lower than reported in other Belgian studies [
12,
18]. In contrast, most studies focus on the hospital physician’s acceptance and not the GPs. In the OPERAM trial, an international study also focusing on optimizing drug treatment in geriatric patients, the approach was similarly to ours, except that the GP acceptance rate was set at two months post-discharge (62%) [
24]. However, the included patients in this study were not surgical patients and their context might have changed more at two months’ post-discharge compared to one month, resulting in a less likely acceptance of interventions. The personal approach of the GP, with telephone contact at two time points, was considered to add to the success of the PIs.
This study underlines that PIs can have a significant impact on preventing drug-related patient harm. However, the pharmacist-led transmural care program is a time-consuming intervention, and selection of patients at highest risk of DRPs is imperative to make it feasible. In this study, a tool to identify older patients at high risk of DRPs, could be useful to target PIs for patients with high pharmaceutical needs due to factors such as illness severity, co-morbidities, high-risk drugs and polypharmacy. These tools should contribute to more informed discussions between patients and GPs on reducing DRPs, thus breaking down some barriers to deprescribing. Literature on how to properly define these high-risk patients is still scarce [
25,
26]. Therefore, more studies are needed on patient selection in order to optimize resources and implement this intensified form of medication review in practice [
2].
Regarding CI, 70% of PIs were rated by the geriatrician as having the potential to avoid patient harm while for the CP it was 45%. These results are similar to other findings [
27]. Our results showed low agreement raters for the inter-rater reliability which is comparable to results obtained in other studies by Somers et al. (Kappa = 0.15–0.25) and lower than Vo et al. (kw = 0.41) suggesting the difficulty of CI assessment [
16,
28]. These differences clearly show the importance of a multidisciplinary approach given the point of view and focus of geriatricians and CPs is different in evaluating the CI of PIs.
The study has potential limitations. First, this study was performed in a single hospital, limiting generalizability. Second, we did not assess the effect of the intervention on clinical outcomes such as readmissions as our sample size would not be large enough to ensure adequate power. Neither did we evaluate other relevant outcomes of resolved DRPs, such as cost savings and QOL, or whether the admission was drug-related. The success rate of a PI may be influenced if the GP is made aware of the fact that the admission was drug-related. Another limitation was that patient findings regarding the pharmacist-led transmural care program were not documented. Only a few studies actively involved the patient, as patient input is of high importance [
29]. If patients agree with the proposed PIs, they are more likely to follow through with the changes and the chance of the intervention being successful is higher [
30]. In this study, the patient was only actively involved in MR and discharge counseling. Finally, the CI of the PIs should be evaluated by an expert panel in order to provide a more balanced result, integrating all perspectives of the involved care providers.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.