Staff training to reduce behavioral and psychiatric symptoms of dementia in nursing home residents: a systematic review of intervention reproducibility

Staff training has been cited as an effective intervention to reduce behavioral and psychiatric symptoms of dementia (BPSD) in nursing home residents. However, the reproducibility of interventions can be a barrier to their dissemination. A systematic review of controlled clinical trials on the effectiveness of staff training for reducing BPSD, published between 1990 and 2013 on the EMBASE, PUBMED, LILACS, PSYCHINFO and CINAHL databases, was carried out to evaluate the reproducibility of these interventions by 3 independent raters. The presence of sufficient description of the intervention in each trial to allow its reproduction elsewhere was evaluated. Descriptive analyses were carried out. Despite reference to a detailed procedures manual in the majority of trials, these manuals were not easily accessible, limiting the replication of studies. The professional expertise requirement for training implementation was not clearly described, although most studies involved trainers with moderate to extensive expertise, further limiting training reproducibility.

B ehavioral and psychiatric symptoms of dementia (BPSD) are highly frequent particularly at moderate to severe stages. 1 These symptoms are very distressing and represent one of the leading causes of institutionalization of demented subjects. 2 It has been estimated that BPSD prevalence in patients with dementia living in nursing homes is around 80%. 3,4 Pharmacological treatment for BPSD has shown poor response, 5 thus non-pharmacological therapies have a place in this scenario. [6][7][8] Systematic reviews have demonstrated that staff training is an effective intervention for reducing BPSD in residents of nursing homes, although methodological weaknesses of trials and the limited number of largescale studies have been highlighted. 6,[8][9][10] Concerns about feasibility and reproducibility of staff training have also been raised. [8][9][10] A new study using an intervention that has already been applied will only prove feasible if sufficient information about the original procedures is provided in previous studies. It is important to ascertain whether interventions were described in such a way that makes them amenable to replication. Therefore, the aim of this study was to evaluate the provision of welldescribed operationalization of staff training programs and the presence of guidelines that ensure their reproducibility in future studies.

methodS
The process for selecting studies was based on the Cochrane Handbook for Systematic Reviews of Interventions. 11 Firstly, a systematic literature search was carried out using five databases: EMBASE, PUBMED, LILACS, PSYCHINFO and CINAHL. Three search strategies employing the following keywords were performed: [1] "staff education" OR "staff training" OR "staff development" OR "nursing staff" OR "nursing" AND "dementia" OR "Alzheimer's disease"; [2] "nursing home" OR "care home" AND "caregiver"; [3] "desenvolvimento de pessoal" OR "recursos humanos de enfermagem" AND "demência" OR "doença de Alzheimer".
In the second step, titles and abstracts retrieved were reviewed according to the inclusion criteria in order to be selected for full-text revision. These inclusion criteria were: [ In the third step, a full-text revision was carried out by three independent raters in order to select the studies. For this step, the same inclusion criteria cited above plus an eighth criterion, i. e., the presence of training effectiveness in reducing BPSD, were employed.
The presence of a detailed training and procedures description was investigated in the final full-text trials included. Training theoretical framework, intensity of training program (session number, duration and frequency and total duration), presence of individual training, presence of a manual describing all aspects of training, and professional expertise requirement for training implementation were evaluated. Descriptive analyses were performed. reSultS included studies. The search strategy identified 1257 studies, comprising 327 duplicated studies (same abstract found twice) initially discarded. Abstract screening resulted in the exclusion of 816 studies. The main reasons for exclusion were being unrelated to the topic (339), intervention other than staff training (176) and study design other than controlled clinical trial (121). Of the 114 full-text studies retrieved for detailed inspection, 97 were excluded for not having fulfilled the inclusion criteria. The study design was the most frequent reason for exclusion (80). Seventeen studies were preliminarily considered for review. One out of the 17 had the same results published twice and was therefore excluded. 12 Additionally, four studies that failed to show staff training effectiveness in reducing BPSD were also excluded, [13][14][15][16] give the main objective was to evaluate the reproducibility of effective interventions. This selection process resulted in 12 studies for final inclusion. Figure 1 contains a detailed diagram of retrieved, excluded and included studies. descriptive analyses of staff training reproducibility.
[a] Theoretical framework of staff training: categories were based on the division proposed by Spector and colleagues, 10 described as: (i) behavioral-oriented approach with person-environment fit; (ii) communication approach; (iii) person-centered approach; (iv) emotion-oriented approach; and (v) other approaches. The behavioraloriented approach with person-environment fit was the most frequent framework applied (4 out of 12 studies). [17][18][19][20] This approach takes into account that behaviors are maintained through reinforcement and considers the necessity of adaptation of the environment to suit individual needs. Two studies 21,22 used person-centered approach (framework focused on individual needs and abilities) and another two trials employed instruction cards to advise on the management of BPSD, showing similar general guidelines but addressing different symptoms. 23,24 A myriad of other theoretical frameworks were found as singular instances, including com-munication skills aimed at teaching staff communications strategies to prevent and deal with behavioral problems; 25 an emotion-oriented approach which helps staff understand and validate the feelings of residents; 26 abilities-focused care, based on the concept of recovery and maintenance of functional abilities 27 and a specific approach which uses techniques to reduce the need for restraint. 28 [b] Intensity of training program: great heterogeneity in the number, frequency and duration of sessions was found. Training programs took from 2 days 25 to 1 year, 21 with a minimum of 7 25,27 to a maximum of 25.5 hours. 23 Total study duration averaged 4 months, requiring around 15 hours of theoretical and practice exercises.
[c] Detailed procedure description and access (manual detailing all aspects of training, availability of the manual and supplementary material): nine trials provided a brief procedural description in their method section and cited a manual with detailed intervention description and procedures to apply the intervention. [17][18][19][20]22,23,25,27,28 Two manuals, 29,30 referred to by Bird et al. and Chenoweth et al.,respectively,19,22 are freely available only to affiliated members of an American University. Other requests for these documents made to the university library are charged. The manual 31 cited by Lichtenberg et al. 17 can be purchased on Amazon's website while a training program with videotapes and a written manual cited by Teri et al. 18 can be purchased on the University of Washington website. 32 Deudon et al. 23 supplied a website address 33 from which to obtain the complete set of instruction cards (in French) used during their teaching program, to provide caregivers with practical information on what to do and how to respond when faced with BPSD, but it was not possible to retrieve the document from the website. Verkaik et al. 20 applied an intervention adapted from staff training used in a previous study, 34 however they did not describe the adaptation procedures. Three trials that had cited manuals detailing procedures 25,27,28 did not describe how to access the material. One trial 21 referred readers to two other studies for a more in-depth description. 35,36 All available manual and supplementary material were published in the English language, except for the set of instruction cards cited by Deudon et al. 23 which was in French. Senior authors from two studies 17,18 expressed their availability to resolve doubts concerning the method or to provide handouts and didactic material.
[d] Professional expertise requirement for training implementation: although not all trials indicated the professional expertise requirement for training implementation, most of the studies (9 out of 12) described the expertise of study trainers involved. Nurses, 19,21,22 psychologists 18,19,24 and nursing assistants 17,20 were the professionals who most frequently had implemented the training. Five studies reported moderate to extensive expertise of the trainer conducting the intervention: one study described that the trainers had "geriatric mental health experience" 18 and in another trial 25 the person who implemented the intervention had "extensive group leadership experience"; trainers from the study of Chenoweth et al. 22 had experience of "hundreds of hours of intervention procedures" before its implementation and the trainers of the research carried out by Deudon et al. 23 were depicted as "professionals with extensive experience of working with residents with dementia".
The characteristics related to the intervention reproducibility of the included studies are summarized in Table 1.

diScuSSion
We carried out a systematic review to evaluate the reproducibility of staff training as an effective intervention for BPSD in nursing home residents. Issues concerning theoretical framework, intensity of training program, presence and availability of detailed procedural description and professional expertise requirement for training implementation were evaluated. We concluded that despite references to detailed procedures manuals in the majority of trials, these manuals were not easily accessi-ble, limiting the replication of the studies. Furthermore, the professional expertise requirement for training implementation was not clearly stated, although the trainers who implemented the programs in most studies had moderate to extensive expertise where this finding also limits training reproducibility. Some important points should be discussed in relation to the presence of detailed description of procedures used in the trials and regarding their accessibility. None of the 12 trials provided sufficient procedure descriptions and none of the manuals cited in the nine studies were accessible free of charge. Additionally, most of the manuals are only published in English language. These facts can hamper intervention replication in countries with different languages and lower socioeconomic levels. Furthermore, each research group applied a specific procedure, although some of them used similar theoretical frameworks. This lack of training standardization prevents broader use and generalization of the interventions.
The training programs were based on approaches ranging from inductive practices ("do this" and "don"t do that") 23,24 to models based on observed behaviors. [17][18][19][20] They also employed practices that require the identification of communication problems, 25 emotional comprehension 26 and personalized care with the elderly. 21,22 Clearly, higher approach complexity demands greater expertise from the trainers. Although it was not clearly specified what professional expertise is required to implement the training, the programs were conducted by professionals with moderate to extensive experience in the theoretical model and training. This aspect should be considered because the more expertise required, the less feasible and replicable the intervention.
In terms of the number of sessions and total time involved in the programs, these could all be replicated. Nevertheless, the large-scale adoption of training programs with higher intensity requires more planning and organization. We did not evaluate whether the requirements of institutional organization for training implementation were cited in the trials and this represents a limitation of our review. The non-consideration of organizational and system factors in long-term care facilities when planning and implementing training initiatives has been cited as one aspect responsible for difficulties in the sustained transfer of knowledge to practice in staff training programs. 37 Behavioral and psychological symptoms of dementia are challenging, distressing and very frequent in nursing home residents. More effective therapies for these symptoms are still needed. Staff training programs appear to be a good option in this scenario, but require more standardization. The extent to which a specific training can be repeated, i.e., its reproducibility, is a crucial requirement to carry out a proper evaluation of its effectiveness. Considering results obtained from this review, we believe that concerted efforts should be made to develop a universal, feasible, standardized and easily accessed training program that can be implemented and evaluated worldwide in different cultures and countries.