The population of individuals who reside in assisted living memory care (ALMC) is increasing. This population presents with responsive behaviors that reflect unmet physical, emotional, and environmental needs, which are characterized by physical or verbal expressions that are labeled by staff as “challenging behaviors.” Although recognized that engaging the individual in meaningful activities can decrease these responsive behaviors, finding meaningful social activities that are reflective of the individual's interests and abilities can be difficult. In a scoping review, Hitzig and Sheppard (2017) noted that Montessori-based programing is an emerging approach for promoting engagement of individuals with dementia.
The Goodman Group initiated the implementation of the Montessori-Inspired Lifestyle® (MIL) in 2015 as a foundation of care for residents in ALMC neighborhoods. The outcomes of the implementation of the MIL were investigated through a quality improvement process.
The MIL is a resident-driven approach in providing care for adults living with dementia or Alzheimer's disease (Camp, 2010). The program, based on the methods of Maria Montessori as adapted for adults by Dr. Cameron Camp (2006), uses Montessori principles that encompass task breakdown, guided repetition, and progression from simple to complex and concrete to abstract. These Montessori principles are integrated with principles used in dementia interventions, including extensive use of external cues and reliance on procedural or implicit memory rather than declarative or explicit memory. Activities and everyday interactions are used as a means to reconnect persons with dementia and create opportunities for engagement within their community (Camp, 2010). As a foundation of care, the MIL was integrated into every interaction with the resident on a 24-hour basis, not just for activities. The implementation of the MIL allowed the team to build on the current program and intensify the focus on discovery of residents' strengths and abilities to determine meaningful activity for each individual.
Implementation of the MIL was sequenced over several years, based on budget and readiness, at various ALMC neighborhoods throughout the United States. All staff who work in an ALMC neighborhood (e.g., nursing, life enrichment, dietary, facilities) were required to participate in training. In addition to this training, changes in the environment were made to support the implementation of the MIL as a model of care. In addition, education on the MIL was provided to non-staff members, such as volunteers and family members, as they are integral to the implementation of the MIL in the neighborhood.
Administrative support was essential for the sustainability of the MIL. Neighborhood teams were provided with multiple resources to support successful implementation over a 12-month period. One approach was a focus on one MIL principle at daily shift meetings, with team members asked to share success stories using that specific principle. External aids were posted to remind staff and guests about the MIL principles. The MIL principles were incorporated into reoccurring events for visitors and recognized informally to help integration on a routine basis. Action plans for 6 months and 12 months were used to support and evaluate the implementation plan. The MIL corporate project lead conducted weekly check-in calls and made quarterly on-site support visits. Regional and national directors incorporated review of the MIL implementation as part of every site visit and consulted with the MIL teams on a regular basis.
Two literature reviews (Hitzig & Sheppard, 2017; Sheppard et al., 2016) provided insight into the approaches that have been used in the implementation of Montessori methods and the quality of the research related to outcomes. In the scoping review of the implementation of Montessori methods for individuals with dementia, four categories of approaches were identified: (a) staff assisted, (b) intergenerational, (c) resident assisted, and (d) volunteer or family assisted (Hitzig & Sheppard, 2017). The approaches were all “activity” or “program” focused and had a high variability in implementation. No standardized protocol or best practice for implementation of a Montessori program was evident.
The systematic review by Sheppard et al. (2016) that addressed the benefits of Montessori-based activities for persons with dementia found varying degrees of evidence-supported outcomes of the programs. Strong evidence was found for benefits on eating performance. The outcomes related to cognition were not as strong but suggested that lower-level abilities, such as memory and attention, may benefit from Montessori-based activities. Improvements in affect and engagement were reported in 10 different studies, although there were mixed levels of evidence related to methodological quality. Montessori programming was found to heighten constructive engagement (e.g., motor or verbal behavior in response to the activity), reduce passive engagement (e.g., display of listening or looking behavior in response to the activity), and promote a more positive affect in persons with dementia. As noted in the systematic review (Sheppard et al., 2016), the types of Montessori activities of the various studies reporting the above-mentioned outcomes represented different methods or approaches. Although encouraging, these results do not address the integration of the Montessori principles into the foundation of care by the entire team, which is the focus of the MIL. Hitzig and Sheppard (2017) state that “adopting the principles underlying MBP [Montessori-based programming] as a philosophy throughout the facility may accelerate the person-centered culture change currently being pursued in LTC [long-term care] and other dementia care settings” (p. 112).
The purpose of the current quality improvement project was to evaluate the MIL as a foundation of care at several ALMC neighborhoods owned and managed by one corporation. Previous studies have focused on the outcomes of Montessori-based programing for specific activities, such as eating, intergenerational experiences, and group and one-on-one experiences. This project provides the outcomes from a neighborhood culture change with the integration of the MIL as a foundation of care.
The Comprehensive Process Model of Engagement (Cohen-Mansfield et al., 2009) guided the development of the project. The model ascribes three categories of attributes as contributing to engagement: (a) environmental, (b) stimuli, and (c) person. The Comprehensive Process Model of Engagement specific to the MIL program addresses each of these attributes. The attributes as defined for this evaluation and the outcomes are presented in Table 1.
Comprehensive Process Model of Engagement Specific to the Montessori-Inspired Lifestyle® (MIL)
Approach for Evaluation
The 2015 Standards for Quality Improvement Reporting (SQUIRE 2.0; access http://squire-statement.org) served as a guide for this quality improvement project due to low risk of harm or burden to participants with potential to improve the quality and safety of the select population.
Engagement of the resident was selected as a primary outcome as it reflects the integration of each component of the MIL. Engagement was defined by Cohen-Mansfield et al. (2009) as the “act of being occupied or involved which is influenced by environment, stimulus, and personal characteristics” (p. 300). If a resident is not engaged in meaningful activities, negative effects occur. Non-engagement magnifies apathy, boredom, depression, and loneliness, which often accompany the progression of dementia (Cohen-Mansfield et al., 2009). Definitions of the attributes of the four levels of engagement are provided in Table 1.
An Engagement Observation Recording Form was developed to measure engagement based on the Montessori principles. The form was adapted from a tool used to measure the engagement of individuals with dementia to various stimuli in the Pearls of Life® Program (Westberg et al., 2017). The initial version of this tool was presented by Cohen-Mansfield et al. (2011) in capturing engagement of individuals with dementia to various stimuli. The tool is now routinely used by staff to capture level of engagement.
Outcomes of the MIL implementation were evaluated using naturally occurring groups. The first group comprised residents of ALMC units where the MIL was implemented prior to the time of the project; thus, an engagement observation was not captured before implementation. The second group comprised residents of ALMC units for which engagement was captured before and after MIL implementation. Follow up was approximately 7 to 12 months after implementation for all but one neighborhood. This neighborhood had a 1-month follow up, as the team was interested in determining changes in the initial implementation.
A subgroup of the second group of residents comprised residents who were observed before and after implementation of the MIL. Although there were only 13 residents in this subgroup, the ability to use matched pairs allowed for a strengthened design of residents as their own control.
Through a previous observation study (Westberg et al., 2017), it was determined that four to five residents could be observed during one 12-hour period to reliably capture engagement. Residents who represented a range of cognitive and physical abilities, as well as current engagement levels, were selected as participants.
The first author (P.M.G.) was responsible for all observations and recordings of engagement, which took place over a period of 2 years. Observation was conducted for 10 minutes for seven activities for each participant. The time period of 10 minutes for capturing level of engagement was supported in the pilot period, as well as by Casey et al. (2014) and Westberg et al. (2017). The seven activities represented four non-meal activities (two planned and two unplanned) and three meals.
Resident characteristics were documented from chart review and included age, gender, Global Deterioration Score (GDS), and admission date to the neighborhood. The GDS is completed to assess cognitive abilities every 6 months or at a change in condition for residents of ALMC neighborhoods. Scale scores range from 1 (no cognitive decline) to 7 (very severe cognitive decline) (Reisberg et al., 1982). Medication orders were reviewed from electronic health records (EHRs) when available.
Data were entered into IBM SPSS version 25 for analysis. Comparisons were conducted to determine differences between individual characteristics (age, gender, and GDS score) for baseline and follow-up groups, with no significant differences found. Change in minutes of positive engagement was used to represent engagement outcome. Differences in minutes of positive engagement before and after implementation of the MIL was conducted using t test analysis.
A total of 85 resident observations were conducted to ascertain the level of engagement during meals and planned and unplanned activities. Of these 85 observations, 72 different residents (53 [73.6%] women and 19 [26.4%] men) were included, 13 of whom were observed twice (i.e., before and after MIL implementation). Mean age of the total group was 86.08 years (range = 70 to 98 years, SD = 7.37). GDS scores ranged from 3 (mild cognitive decline) to 7 (very severe cognitive decline), with >60% of participants scoring 5 or 6, reflecting moderate dementia. Average length of time from admission to time of observation was 1 year 9 months (range = <1 month to 7 years).
The total group (N = 85) was categorized by time of observation, before (n = 33) or after (n = 52) MIL implementation. There were no differences in age or GDS scores between groups. The differences in minutes of positive engagement before and after implementation of the MIL for meals and planned and unplanned activities increased for all three activity types and were statistically significant (Table 2).
Positive Engagement Minutes Before and After Montessori-Inspired Lifestyle®
Engagement observations before and after MIL implementation were possible in six of the nine neighborhoods. Twenty-five participants were observed before MIL implementation and 28 participants were observed after MIL implementation. As presented in Table 2, there was an increase in the positive engagement minutes for all meals and planned and unplanned activities, with the increase in planned activities being statistically significant (t = 2.51; p = 0.015).
Thirteen residents comprised a subgroup of participants in the neighborhoods that had a before and follow-up observation. Mean age of these residents was 83.08 years (range = 73 to 98 years), which was slightly younger than the overall group. The gender distribution was similar to the total group, with three men and 10 women. Of nine residents who had a GDS score, four residents scored a 5, and five residents scored a 6. Average GDS scores of this group are similar to the overall group, yet the higher-level categories of cognitive ability (GDS score = 3 or 4) are not represented in this subgroup. A paired t test was used to determine whether a significant difference existed in the positive engagement minutes before implementation of the MIL to after implementation for participants in this subgroup. There was no significant difference in positive engagement for any of the activities (meals, planned and unplanned), but mean positive engagement was higher after implementation for planned and unplanned activities. There were little positive engagement minutes for meals for this subgroup before and after MIL implementation (Table 2).
Further analysis evaluated the effects of gender, cognitive ability, and setting on the minutes of positive engagement. There were no significant differences in minutes relative to gender for any of the three activity types. There were significant differences in cognitive ability and setting related to positive engagement minutes. GDS scores were categorized into three groups: Group 1, scores of 3 and 4 (mild impairment); Group 2, score of 5 (moderate impairment); and Group 3, scores of ≥6 (severe impairment). Residents in Group 1 (mild impairment) had the highest positive engagement minutes for all three activity types. Residents in Group 2 (moderate impairment) had the lowest minutes for meals and unplanned activities, with those in Group 3 (severe impairment) having the lowest positive minutes in planned activities. The analysis also indicated that there were differences in settings, with a post hoc analysis identifying the difference between two sites: the highest minutes of positive engagement were in the neighborhood with residents with the highest level of cognitive ability and the lowest positive engagement minutes were in the neighborhood with residents who had the most severe impairment in cognitive ability. (It is noted that admission to some neighborhoods is based on best-fit related to cognitive ability.)
The time period of exposure to the MIL environment for residents who were observed was calculated for the time of MIL implementation, time of admission to the neighborhood, and date of data collection. Of the 52 residents who had been exposed to the MIL, length of exposure varied from 1 month to 18 months (mean = 6.96 months; SD = 5.53 months). There was a statistically significant negative correlation between length of exposure and minutes of positive engagement for meals (r2 = −0.356; p = 0.01) and unplanned activities (r2 = −0.27; p = 0.05), with planned activity minutes (r2 = −0.316; p = 0.02) approaching significance. The correlation was in the negative direction, indicating that as time of exposure to the MIL increased, the minutes of positive engagement decreased.
Changes in the prescribing of as-needed (PRN) psychotropic medications were explored as an indicator of change in behavior problems of participants. The absence of a PRN medication was considered reflective of adequate control of behaviors with routine medications and the nonpharmacological management with MIL implementation.
Medication records were available for 53 participants. A graphic comparison of the scheduling categories of psychotropic medications before and after MIL implementation was conducted. Several positive clinical changes were noted: (a) participants not prescribed psychotropic medications (routine or PRN) increased from six to seven; (b) one participant who was taking a PRN psychotropic medication had this medication discontinued; and (c) the number of participants on both a routine and PRN psychotropic medication decreased from 11 to nine. Although these changes are minimal, the effort to decrease use of PRN psychotropic medications among this group is important clinically.
The overall increase in minutes of positive engagement following implementation of the MIL is encouraging. Total group minutes of positive engagement was higher for all three types of activities at a statistically significant level for residents who were observed following MIL implementation compared to residents who were observed prior to MIL implementation. There was also an increase in positive minutes in all three types of activities when comparing the groups of residents in the same neighborhood before and after MIL implementation. The increase in positive engagement minutes was only significant for planned activities. An increase in positive engagement for Montessori-based programs was also reported by Sheppard et al. (2016) as an outcome of multiple studies.
The subgroup of 13 residents who had before and after observations allowed for analysis with these residents as their own control. This group had an increase in positive engagement minutes for planned and unplanned activities, with minutes during meals very low and with no change. The small number of residents in this subgroup was a limitation. It is also recognized that four residents in this group were those for whom follow up was conducted 1 month after implementation.
Residents with the highest levels of cognitive ability were the most engaged. This finding supports the work of Mahendra et al. (2011) who indicated that Montessori-based programs are best suited to those with the ability to communicate verbally and/or understand task instructions. Residents with moderate cognitive ability had the lowest positive minutes in engagement at meals and unplanned activities. These individuals may be struggling with the challenges of maintaining their independence, but staff are not recognizing them as needing more prompts, physical and verbal, for engagement. Individuals with more severe cognitive impairments are recognized by staff as needing more prompts for engagement. As noted previously, the difference in positive minutes in engagement in the three types of activities by setting probably reflects the cognitive ability of residents. Several neighborhoods are in the same facility, with admissions to specific neighborhoods based on the cognitive ability of the resident.
An additional finding of the evaluation was decreased orders of PRN psychotropic medications, although slight, among residents. Clinically, a PRN psychotropic medication is ordered when there are abrupt changes in a resident's behavior that are considered challenging by staff. The decrease in PRN medication is thought to reflect effectiveness in management of behaviors through the use of routine psychotropic medications along with the use of nonpharmacological management, which is incorporated in the MIL.
There are limitations to the current study. Being a quality improvement project, the results are not generalizable. The residents observed were purposefully selected to represent various levels of cognitive and physical abilities, specifically including those residents who staff identified as presenting with challenging behaviors. In an attempt to have residents serve as their own control, some residents were observed before and after MIL implementation. Yet, this group was small, as a number of residents had significant changes in condition that resulted in transfer or death. It is recognized that residents with GDS scores that reflect severe cognitive impairment are nearing end of life, contributing to the small numbers in this subgroup.
The time of observation was only a total of 70 minutes (10 minutes × seven activities) during 1 day, which is a limitation of the evaluation. The times were randomly selected during meals and planned activities. Unplanned times were based on residents being in common areas when no planned activities were occurring. The activities residents participated in also varied based on the schedule for the day; however, categories of activities did not differ. Although categories were similar before and after MIL implementation, it is recognized that with the MIL, residents determine the events for the day. There is also a difference in the approach to each activity based on the MIL principles, which resulted in higher levels of positive engagement.
The negative correlation that was found between the time period that a resident was exposed to the MIL and minutes of positive engagement identifies the need to continue training of staff in MIL principles and the need to develop a process to ensure sustainability of the MIL. Sustainability is a challenge for implementation of various interventions and, even with the multiple approaches mentioned previously, the integration of the MIL is a continual focus.
Engagement of individuals with dementia in ALMC is important in decreasing responsive behaviors that reflect unmet physical, emotional, and environmental needs. The results of this quality improvement project indicate that implementation of the MIL as a foundation of care increases engagement of ALMC residents. Resources to support the implementation and sustainability of the MIL and a commitment of the team to the MIL are essential to create the opportunity for the highest level of engagement of residents.
The three most common unmet needs in individuals with dementia have been identified as boredom/sensory deprivation, loneliness/need for social interaction, and need for meaningful activity (Cohen-Mansfield et al., 2015). The current quality improvement project has indicated that the MIL as a foundation of care for residents of ALMC neighborhoods increased positive engagement minutes throughout the day during meals and planned and unplanned activities. Outcomes of this project have also raised the awareness of the team in creating meaningful engagement opportunities with consideration of diverse cognitive abilities as well as recognition of the need for reoccurring training for the sustainability of this model of care.
- Camp, C. J. (2006). Montessori-based dementia programming in long-term care: A case study of disseminating and intervention for persons with dementia. In Hyer, L. & Intrieri, R. (Eds.), Geropsychological interventions in long-term care (pp. 295–314). Springer.
- Camp, C. J. (2010). Origins of Montessori programming for dementia. Non-Pharmacological Therapies in Dementia, 1(2), 163–174 PMID:23515663
- Casey, A. N., Low, L. F., Goodenough, B., Fletcher, J. & Brodaty, H. (2014). Computer-assisted direct observation of behavioral agitation, engagement, and affect in long-term care residents. Journal of the American Medical Directors Association, 15(7), 514–520 doi:10.1016/j.jamda.2014.03.006 [CrossRef] PMID:24755476
- Cohen-Mansfield, J., Dakheel-Ali, M. & Marx, M. S. (2009). Engagement in persons with dementia: The concept and its measurement. The American Journal of Geriatric Psychiatry, 17(4), 299–307 doi:10.1097/JGP.0b013e31818f3a52 [CrossRef] PMID:19307858
- Cohen-Mansfield, J., Dakheel-Ali, M., Marx, M. S., Thein, K. & Regier, N. G. (2015). Which unmet needs contribute to behavior problems in persons with advanced dementia?Psychiatry Research, 228(1), 59–64 doi:10.1016/j.psychres.2015.03.043 [CrossRef] PMID:25933478
- Cohen-Mansfield, J., Marx, M. S., Freedman, L. S., Murad, H., Regier, N. G., Thein, K. & Dakheel-Ali, M. (2011). The comprehensive process model of engagement. The American Journal of Geriatric Psychiatry, 19(10), 859–870 doi:10.1097/JGP.0b013e318202bf5b [CrossRef] PMID:21946802
- Hitzig, S. L. & Sheppard, C. L. (2017). Implementing Montessori methods for dementia: A scoping review. The Gerontologist, 57(5), e94–e114 doi:10.1093/geront/gnw147 [CrossRef] PMID:28062518
- Mahendra, N., Scullion, A. & Hamerschlag, C. (2011). Cognitive-linguistic interventions for persons with dementia. Topics in Geriatric Rehabilitation, 27, 278–288 doi:10.1097/TGR.0b013e31821e5945 [CrossRef]
- Reisberg, B., Ferris, S. H., de Leon, M. J. & Crook, T. (1982). The Global Deterioration Scale for assessment of primary degenerative dementia. The American Journal of Psychiatry, 139, 1136–1139 doi:10.1176/ajp.139.9.1136 [CrossRef] PMID:7114305
- Sheppard, C. L., McArthur, C. & Hitzig, S. L. (2016). A systematic review of Montessori-based activities for persons with dementia. Journal of the American Medical Directors Association, 17(2), 117–122 doi:10.1016/j.jamda.2015.10.006 [CrossRef] PMID:26589532
- Westberg, K., Gaspar, P. M. & Schein, C. (2017). Engagement of residents of assisted living and skilled nursing facility memory care units. Activities, Adaptation and Aging, 41(4), 330–346 doi:10.1080/01924788.2017.1376175 [CrossRef]
Comprehensive Process Model of Engagement Specific to the Montessori-Inspired Lifestyle® (MIL)
| MIL||MIL implemented|
| Non-MIL||MIL not implemented|
| Meals||Routine dining experiences: breakfast, lunch, and dinner (if one of the meals was not attended, a snack was substituted)|
| Planned activities (e.g., reading group, cooking, music program, music group)||Scheduled event on the event calendar|
| Unplanned activities (e.g., 1:1 or group discussion, individual table activity)||Non-scheduled events that occur in common areas of the unit|
| Cognitive ability||Global Deterioration Scale score|
| Prescribed psychoactive medications||Prescribed routine and as-needed psychoactive medications|
| Age||Age in years|
| Engagement attribute categories|
| Positive engagementa||Smiles, laughs, or shows other outward manifestation of happiness, such as hugging/physical touch or active nodding. Reaching out or using manipulatives. Focused on activity—limited distraction to outside stimulus. Attempts conversation and/or is actively engaged in conversation.|
| Neutral engagement||Eyes track movement of the activity, no change of facial expression, limited physical movement or engagement, little to no attempt at conversation or speaking. Does not use manipulatives.|
| Negative engagement||Assertively pushes stimulus away, cursing, yelling, manifesting frustration at the activity or other entity, other manifestations of negativity, actively walking or wheeling away from stimulus.|
| Failed to engage||All strategies used were unsuccessful. Attempts to engage did not produce any observable results.|
|Engagement duration (time per engagement category)|
| Amount of time resident shows being physically and cognitively responsive to the activity/stimulus||Duration of engagement measured during the 10 minutes of observation for each category of engagement|
Positive Engagement Minutes Before and After Montessori-Inspired Lifestyle®
|Group/Activity||Time Period||n||Mean||SD||SEM||t||df||p Value|
| Planned activity||Before||33||3.37||3.04||0.53||−2.33*||80.33||0.02|
| Unplanned activity||Before||33||2.14||2.81||0.49||−2.25||83.00||0.03|
|Same Unit Comparison Subgroup|
| Planned activity||Before||25||2.81||3.02||0.60||−2.51*||48.54||0.015|
| Unplanned activity||Before||25||2.03||2.58||0.52||−1.67||51||0.102|
|Same Resident Paired Subgroup|
| Planned activity||Before||13||2.56||2.66||0.74||−2.04||12||0.064|
| Unplanned activity||Before||13||2.29||3.15||0.87||−1.41||12||0.183|