Implementation of the acutely presenting older patient (APOP) screening program in routine emergency department care

Older patients form an increasing proportion of emergency department (ED) admissions worldwide and are at higher risk of adverse health outcomes compared to younger patients [1]. The presence of multiple comorbidities, cognitive disorders and atypical disease presentations requires more staff time and resources [2], increases ED length of stay (LOS) and poses organizational challenges [3, 4]. A comprehensive geriatric assessment (CGA) is an effective method to improve older patients’ outcomes [5] but CGA is time-consuming and therefore cannot be routinely applied to every older patient attending the ED. Alternatively, a two-step approach can be used with identification of patients with the highest risk of adverse outcome as a first step, followed by targeted interventions according to the principles of CGA [6, 7]. To this end, several screening instruments and interventions have been specifically developed for older patients in the ED [8, 9] yet few have successfully been disseminated in clinical ED practice.

The acutely presenting older patient (APOP) screening program consists of screening with the APOP screener followed by interventions aimed to improve overall ED care and follow-up of older patients [10]. The program was implemented in routine ED care in the Leiden University Medical Center (LUMC) together with an education program to enhance awareness amongst nurses and doctors working in the ED. There is extensive evidence that effective implementation of complex interventions can be associated with better outcomes in various settings outside the ED, which implicates that evaluation of implementation is an absolute necessity in program evaluation [11, 12]. One of the important reasons why screening of older ED patients is rarely carried out in routine care, is the fact that little is known about the practical issues and feasibility of implementation in everyday ED practice [13], although it was recently shown that administration of the APOP screener is feasible in routine ED practice [14].

In the present study we aimed to evaluate the effects of implementation of the APOP screening program in routine ED care by assessing the compliance with interventions in the ED, during hospital admission and after discharge, and the impact on process of care measures, shortly after implementation. We hypothesized that the implementation of the screening program would not negatively influence the usual ED process, for example no prolongation of the ED stay and it would result in improvement of the care for older patients, for example the increase in geriatric assessments.

During the 2‑month observation period before implementation (“before”) 4614 patients visited the ED of which 920 (20%) were patients aged ≥ 70 years who were triaged at ED arrival. In the 2‑month observation period after implementation (“after”) 953 out of 5188 (18%) ED patients were triaged patients aged ≥ 70 years. Of all triaged older patients, 62% (N = 574) was observed “before” and 59% (N = 560) “after” in order to evaluate the compliance with interventions (Fig. 3).

In this study, the first effects of implementation of the APOP screening program in routine ED care were evaluated after 1 month by assessing the compliance with interventions and the impact on process of care measures. Interventions for high-risk patients in the ED were partly adhered to. Implementation of the program resulted in increased numbers of executed CGAs during hospitalization, communication of screening results to the GP and telephone follow-up after ED discharge. Implementation had no major effects on ED LOS and hospital admission.

To the best of our knowledge this is the first study evaluating the implementation of a multicomponent screening program for older patients comprising screening and targeted interventions in routine ED care. In a recent substudy we showed that implementation of the APOP screener was feasible with a screening rate of 59% [14]. Compared to other studies [13, 24, 25] our screening rate assessed shortly after implementation in routine ED care is relatively high. A screening rate of 100% is difficult to achieve because the time restraints inherent to a busy ED will prevent nurses administering the screener. Since there are only few ED multicomponent studies published [26] we are only able to compare single components. In one study, telephone follow-up for all older ED patients resulted in 97% successfully contacted patients of which 40% required clarification of home care instructions [27], comparable to our results in high-risk older patients. The use of a clinical risk prediction tool to select high-risk patients and target interventions to those patients most likely to benefit, the increased proportion of patients who receive CGA and the improved communication of screening results to the GP have been associated with improved patient outcomes in other settings [8, 9, 28]. Definitive proof of (cost)effectiveness of the APOP screening program on patient outcomes, such as functional decline, should come from future studies, for example by using a multicenter stepped-wedge design [29].

The present study has several important findings for clinical practice. Firstly, implementation of screening in the ED resulted in improved execution of some individual interventions for older patients during their ED stay, i.e. adequate nutrition. However, the intervention “presence of family” did not increase, probably because this was already very high before implementation, i.e. a ceiling effect. The interventions “nursed on bed” and “room with daylight” also did not improve, probably because they were less feasible due to a lack of capacity (in our ED there are few beds and rooms with daylight available). Secondly, program implementation resulted in a significant increase in the number of executed CGAs, which has been shown to be an effective method to improve outcomes [5]. In 71% of the high-risk patients CGAs were executed during hospitalization. Therefore, although interventions in the ED are not always executed, screening is a useful first step to ensure that high-risk patients receive optimal care during hospitalization. The same holds for high-risk patients discharged home from the ED, of which 79% were reached for telephone follow-up. Finally, implementation of our screening program did not lead to prolonged ED LOS or more hospital admissions. After adjustment for the small differences in the before and after group, there even seem to be less hospital admissions after which is important because impact on capacity is relevant to the feasibility and sustainability of the program.

The repetitive use of the PDSA model as a framework for our implementation strategy helped in understanding barriers and facilitators of implementation [14]. Continuation of future PDSA cycles can help to further improve compliance in our ED and can also help others to start implementation of this screening program elsewhere. The results of the present study are therefore the starting point for new evaluation cycles of the program. Until now, we mainly focused our implementation strategy on the ED nurses, the executors of the screening, which also resulted in mainly nurse-led interventions for high-risk patients. In future, we aim to focus more on physicians and use additional education to increase their awareness and promote a more holistic clinical assessment of older ED patients. Moreover, the interventions of our program were based on recommendations from international guidelines and quality indicators [6, 20] and could be updated according to recent recommendations [30]. If other EDs would like to implement a screening program for older patients they can learn from our limitations and adjust their expectations accordingly, i.e. ensure the presence of rooms with daylight and focus on adequate nutrition during an ED stay.

Our study has several strengths. Firstly, to the best of our knowledge this is the first implementation study evaluating screening and interventions for older patients in routine ED care on a large scale, using real-time observations. Secondly, our implementation strategy was guided by the generally used PDSA model for quality improvement, resulting in good understanding of barriers and facilitators of implementation. Lastly, the screening program was implemented and evaluated in an unselected population of older ED patients, which is therefore generalizable to other ED populations.

Our study also has several limitations. Firstly, the before-after study design has time and seasonal variation as a limitation; however, there were no contextual changes between the two data collection periods. Also, we could not detect substantial differences in patient characteristics between the “before” and “after” group. Furthermore, the main outcome measures for the evaluation of the program were process measures—the proportion of hospitalized patients with geriatric assessment and the proportion of discharged patients with follow-up telephone calls—which are likely unaffected by time period or seasonal variation. Secondly, before implementation older patients could not be screened. Therefore, we could only compare compliance with interventions on the level of total group ED patients ≥ 70 years in the before and after periods. Small improvements in compliance with interventions in high-risk patients might therefore have been missed. Finally, the program was implemented in one tertiary care center which limits generalizability.

In conclusion, implementation of the APOP screening program in routine ED care did not negatively impact the ED process and resulted in an increase of CGA and telephone follow-up in older patients. Since this was a first evaluation shortly after implementation, future studies should investigate whether sustainable changes in management and patient outcomes occur after more PDSA cycles.