For depression assessment, a stepped care approach involving three steps is advised. In the first step, “detection,” a nursing staff member uses a screening instrument to recognize residents who may have depressive symptoms. An observer-rated instrument focusing on nonverbal symptoms is advised for this step, since this is not dependent on communication abilities which can be impaired in residents with dementia. To this end, the five-item Nijmegen Observer-Rated Depression (NORD) scale (see gray box) was developed and validated, which can be used for screening for depression on a regular basis in NH residents with and without dementia. It appeared sensitive, short, and easy to score (Leontjevas et al. 2012b).

In the Netherlands, where psychologists are staff members in most care organizations, the second assessment step, “extensive screening,” is to be conducted by a psychologist when this is indicated by the detection instrument used in the first step. For residents who cannot be interviewed because of the severity or type of their (cognitive) impairments, the use of a proxy-based depression screening instrument is advised. For residents who can self-report, the Geriatric Depression Scale 8-item version is suggested ((Jongenelis et al. 2007) GDS8, cutoff score > 2). When observation is more appropriate, the Cornell Scale for Depression in Dementia ((Alexopoulos et al. 1988) CSDD, cutoff score > 7) has acceptable accuracy when professional caregivers are the only source of information, both in residents with and without dementia (Leontjevas et al. 2012a).

When indicated by the second step, in the third step of depression assessment, a psychologist and/or a physician undertake a diagnostic procedure. AiD advises the use of the Provisional Diagnostic Criteria for Depression of Alzheimer’s Disease for residents with dementia (Olin et al. 2002), instead of the Diagnostic and Statistical Manual of Mental Disorders (APA 2000).

Usually, the second and third assessment steps are combined in one session with the residents and/or nursing staff member. In countries where psychologists and/or physicians are not affiliated to the NH, this is even more advisable.

As described, there is limited evidence for the effects of antidepressants in NHs, especially in residents with dementia. Therefore, the focus should be on psychosocial treatment of depression. AiD emphasizes psychosocial treatment, and the revealed effects of the care program underpin that a multidisciplinary approach is advisable (Leontjevas et al. 2013a, b). AiD has three treatment modules that also reflect a stepped care approach.

Module 1 is to be performed for NH residents with depressive symptoms or depression and comprises environmental and behavioral strategies by nursing or care staff and/or an activity therapist. They develop and execute a day structure program (DSP) and a pleasant activities plan (PAP) (Teri et al. 1997; Verkaik et al. 2011). A DSP structures the way days are spent, thereby preventing disruptions of the circadian rhythm and realizing meaningful ways of spending residents’ days. PAP provides the content of the DSP and focuses on involving residents in activities that they find pleasant and also avoiding activities they find unpleasant. A PAP describes goals and ways to realize them.

Module 2 regards treatment by a psychologist and is complementary to Module 1 for NH residents who meet criteria for a depression diagnosis. If the communicative and cognitive abilities allow for talk therapy, this is initiated by the psychologist. The AiD program does not prescribe a specific therapy, but the manual contains precious memories therapy (Bohlmeijer et al. 2010) which is a form of life review therapy adapted for NH residents. If talk therapy is not possible, mediative therapy – in which the psychologist intervenes in the resident’s social environment through educating and coaching the care team – is advised (Hamer 2003).

Module 3 regards pharmacological treatment. If treatment according to Modules 1 and 2 is not effective or if depression is very severe, the physician may consider prescribing antidepressants. For this, the NICE guideline dementia, which summarizes the most recent evidence and recommendations for treatment of depression in dementia, can be applied (https://​www.​nice.​org.​uk/​guidance/​cg42/​evidence/​full-guideline-including-appendices-17-195023341). It is important to consider that, in patients with dementia, there is no evidence in favor of pharmacological treatment of depressive symptoms and minor depression: the known trials are all negative, while research into effectiveness of antidepressants for major depression in dementia stays inconclusive. Therefore, pharmacological treatment should be prescribed with caution and, preferably, when other treatments have appeared to be ineffective. For residents with dementia, when the observed burden of depression is high, AiD suggests the prescription of antidepressants (selective serotonin reuptake inhibitors: sertraline, fluoxetine, citalopram) in combination with non-pharmacological strategies described above. Antidepressants with anticholinergic side effects should preferably be avoided. Effect and side effects (gastrointestinal and central nervous system related, dry mouth, dizziness/falls) should be monitored carefully. For severe major depression (with psychotic symptoms, suicidal thoughts, or refusal to eat), referral to a psychiatric hospital should be discussed with representatives of the resident. In these cases, electroconvulsive therapy may also be considered (Pellegrino et al. 2013).

The AiD trial showed that on dementia special care units, Module 1 was most effective and that on units for somatically frail residents Module 2 was. No evidence was observed for effects of Module 3, pharmacological treatment. In contrast, indications were found that it might have increased apathy (Leontjevas et al. 2013b).

Element 3 has two goals: (1) timely detection of depressive complaints and (2) monitoring and evaluation of treatment in order to adapt it if necessary. If the element “assessment” shows no depressive complaints, AiD proposes to administer the observational scale of step 1 6 months later. Thus, through element 3, AiD is cyclic. This way, the care team remains alert regarding depressive symptoms. If there appear to be depressive complaints, further steps are necessary. If treatment is prescribed, a psychologist and/or physician make a treatment plan in accordance with element 2 and determine how and when to evaluate it. Ideally, this multidisciplinary evaluation takes place within 3 months after having set a diagnosis. As input for the evaluation, administering the assessment instrument of step 2 “screening” is advised as well as checking the diagnostic criteria of step 3 “diagnosing.” During treatment, it is important to monitor both effects and feasibility of the proposed treatment options. All involved professionals have to be attentive to procedural hurdles and side effects when implementing treatment strategies and are responsible for informing each other and for trying to solve the encountered problems.

It is also advisable to monitor apathy when managing depression. Apathy is characterized by diminished motivation in combination with a lack of goal-directed behavior, goal-directed cognition, and emotional affect, which leads to reduced interest and participation in the main activities of daily living, diminished initiative, early withdrawal from initiated activities, indifference, and flattening of affect (Marin et al. 1991; Starkstein et al. 2001; Robert et al. 2009). In NH residents with dementia, apathy is very common (Zuidema et al. 2009; Wetzels et al. 2010; Selbaek et al. 2013; van Reekum et al. 2005; Cipriani et al. 2014; Zwijsen et al. 2014). However, it is rarely considered a problem in that setting (Leone et al. 2013; Zwijsen et al. 2014), even though it has been found to be associated with a number of adverse outcomes, such as high caregiver burden, poor treatment response, reliance on caregivers to initiate activities of daily living, low quality of life, and more rapid cognitive and functional decline (van Reekum et al. 2005; Tagariello et al. 2009; Starkstein et al. 2006; Bakker et al. 2013; Mulders et al. 2016). The overlap of apathy with depression is evident, as loss of interest or pleasure is a principal symptom to diagnose depression, even when depressed mood is not present (APA 2000; Olin et al. 2002). However, various studies support the concept of apathy as a behavioral syndrome that can be discriminated from depression (Levy et al. 1998; Starkstein et al. 2008; Ishii et al. 2009). Recent studies show a combination of overlapping and different risk factors for apathy and depression after stroke (Withall et al. 2011; Yang et al. 2013). To be able to differ between depression and apathy, the diagnostic criteria for apathy in Alzheimer’s disease and other neuropsychiatric disorders (DCA) (Barca et al. 2010) can be added to a diagnostic procedure for depression. Furthermore, the abbreviated Apathy Evaluation Scale (AES-10) was validated in the AiD trial (Leontjevas et al. 2012c) and can be administered next to screening instruments for depression.

Regarding treatment, as described above, the behavioral strategies applied in AiD activate not only residents with depressive features but can also be beneficial for apathy (Brodaty and Burns 2011), whereas pharmacological treatment of depression using antidepressants may induce apathy (Barnhart et al. 2004; Settle 1998). Thus, for monitoring and evaluation of treatment, measuring apathy – for instance, by using the AES-10 – is advised.

AiD was shown to have effects on depression, quality of life, and apathy. Moreover, it was considered relevant and feasible by the involved stakeholders (unit managers, physicians, and psychologists) (Leontjevas et al. 2012d). However, implementation of AiD during the AiD trial was not highly successful. For instance, treatment was not started in more than half of the indicated cases, and, while a DSP and PAP (Module 1) were applied more often in dementia units than in somatic units, treatment by a psychologist (Module 2) was executed almost three times less often in dementia units (Leontjevas et al. 2012d). Furthermore, element 3 (evaluation) was the least often implemented part of AiD in both unit types.

The reasons for suboptimal implementation reported by stakeholders shed light on requirements for adequate depression management. The stakeholders of the AiD trial mentioned several causes of suboptimal implementation in daily care, such as high workload, lack of time, difficulties in multidisciplinary collaboration or within the care team, staff changes and unfulfilled positions, low involvement of management, and reorganizations (Leontjevas 2012). High turnover of staff in the units not only resulted to additional expenditures but also compromised performing the program as planned. In 33 units that participated in the AiD trial, 22 physicians, 19 psychologists, and 11 unit managers changed their job during the intervention period (12 months on average). In 12 out of 33 units, more than 20% of the nursing staff that were educated by the researchers left the unit. It is not surprising, then, that the NH professionals mentioned frequent staff changes and shortage of (qualified) staff as an important barrier to carrying out the program. Supporting stable multidisciplinary staff and empowering personnel appear crucial for introducing care innovations. According to the professionals, implementation would have been facilitated if there were, in addition to less staff turnover, more investment in educational strategies and in activities for residents.

Furthermore, the suboptimal implementation of the evaluation element during the AiD trial calls for additional attention. When treatment strategies are not evaluated, depression may persist due to ineffective strategies that are not fine-tuned. Furthermore, when implementation is not evaluated, barriers to implementation may persist and prohibit the necessary changes. Both require a solid working routine with structurally planned evaluation meetings and room for discussion and treatment optimization.