Illness Management and Recovery (IMR) is a standardized evidence-based psychosocial intervention aimed to help users with severe mental illness (SMI) to achieve a meaningful recovery (Mueser et al. 2006). Grounded in the stress-vulnerability model, it was developed in the early 2000’s as part of the “National Implementing Evidence-Based Practices Project”, a program sponsored by the Substance Abuse and Mental Health Services Administration to facilitate the adoption of evidence-based practices in clinical practice (Torrey et al. 2005). The program combines in a single framework five types of psychosocial interventions (Salyers et al. 2009): psychoeducation to improve knowledge of mental illness, relapse prevention to reduce relapses and rehospitalization, behavioral training to improve medication adherence, coping skills training to reduce the severity and distress of persistent symptoms, and social training to strengthen social support (Mueser et al. 2002). Providers teach these strategies through a combination of educational, motivational, and cognitive-behavioral techniques.

In its more current version, the IMR program is organized into 11 modules (Gingerich et al. 2011) delivered in individual sessions or in groups on a weekly basis. The program requires three to ten months to be completed. After each session, homework is assigned. Participation of family members and caregivers is also highly encouraged (Roosenschoon et al. 2016). Since its inception, IMR has been widely implemented around the globe within different socio-cultural contexts, healthcare systems, and types of mental health services (McGuire et al. 2014; Mueser et al. 2006). It has been mainly implemented in a single setting, either outpatient or inpatient unit, and delivered to selected patients, mostly those with SMI (schizophrenia, bipolar disorder, major depressive disorder, and schizoaffective disorder). IMR has proven to be effective for improving knowledge on psychiatric disorders, reducing relapses, hospitalizations and anxiety secondary to symptoms. Patients who received the IMR intervention became more constant and consistent with their drug compliance (McGuire et al. 2014). The vast majority of published studies have been controlled, with strict inclusion criteria for age, diagnosis, duration for illness, and included a measurement of fidelity and outcomes (Dalum et al. 2018; Roosenschoon et al. 2016; Salyers et al. 2011). Little is known about the implementation of IMR in a real-world setting (Ganju 2003), beyond the boundaries of clinical research (Drake et al. 2001).

An average department of mental health in Italy offers mental health care to a wide array of patients, with different mental disorders, duration of illness, stages of recovery, and settings (inpatient, outpatient, rehabilitation units). The standard treatment is often individual, focused on management of acute episodes, and maintenance of clinical recovery (mainly symptoms reduction); less often the goal is to pursue a personal recovery, and rarely includes psychoeducational tools to increase disease awareness and promote personal mastery in dealing with the illness. Given the published literature on the versatility of IMR in different contexts, it would be of foremost importance to implement IMR within an entire department, with the goal not only to deliver an evidence-based psychosocial intervention to outpatients, but also as a mean to provide a theoretical and practical recovery-oriented therapeutic framework to different professionals operating within the department, regardless their training or expertise.

Thus, the aim of this study is to describe the IMR implementation within a Department of Mental Health in Ferrara, northern Italy, and the preliminary results achieved, in terms of both implementation and users’ outcomes.


Study Design

This is a retrospective study regarding the feasibility of the implementation of the IMR Intervention in the Integrated Department of Mental Health and Pathological Addictions in Ferrara, Italy.

The current study describes the results regarding those who completed the IMR intervention from April 2017 to July 2021 in outpatient services.


The Ferrara province covers a catchment of 2630 square kilometers and 342,000 inhabitants. The Integrated Department of Mental Health and Pathological Addictions (DAISM-DP) offers mental health care to the entire province under a universal healthcare system. The DAISM-DP includes five outpatients Community Mental Health Centers (CMHC) for adults (Cento, Copparo, Codigoro, Ferrara, Portomaggiore), a 15-bed inpatient acute unit, a 15-bed inpatient psychiatric unit for voluntary only admissions, two intensive residential units, two prolonged treatment units, and two day-treatment facilities. The DAISM-DP also includes five outpatient Services for Addictions and Dependence (SerD), and an outpatient Program for Eating Disorders (EDP).

In 2016, the IMR was included in the Continuing Education Plan for the DAISM-DP.

In December 2019, the IMR intervention implementation was included in the Organizational Manual of the DAISM-DP. It was also shared with the Strategic Direction of the Health Agency of Ferrara and included in the General Protocol n.0079622 of the Azienda USL of Ferrara as a key component of the services provided by the DAISM-DP.


Patients’ enrollment started in April 2017 initially within the community outpatient services (CMHC) only, later it was offered to inpatients as well. The case manager, whose role is to assist patients towards their personal and functional recovery, was instructed to offering the IMR intervention to their eligible assigned users. Participants were invited on a consecutive and voluntary basis, during a regular follow-up visit at the CMHC, if they had all the following characteristics: (1) one of the following diagnoses: schizophrenia, delusional disorder, bipolar disorder, major depressive disorder, substance addiction; eating disorder; (2) willingness to attend the IMR sessions; (3) having a current active therapeutic program within the DAISM-DP; (4) being in a stable phase of the illness. Patients were excluded if they were deemed too be too symptomatic (e.g., acute psychosis, severe cognitive impairment, intoxicated) to attend the sessions, or refused the IMR intervention. By the time of this report (July 2021), only those groups conducted in outpatient services were concluded; IMR implementation within the acute inpatient unit was still ongoing and will not be described in this study.



In order to fulfill the above-mentioned strategic program, in 2016 the DAISM-DP first arranged a 12-h IMR training for 60 mental health professionals including psychiatrist, toxicologist, mental health nurse, educators, psychiatric rehabilitation therapist, psychotherapist, social worker. Moreover, in 2020, 14 providers attended a 3-day intensive training with the IMR creators.

IMR Implementation

IMR was delivered, by default, in groups of maximum eight users. The criteria chosen for the composition of the groups were: similar age, history of substance abuse, cognitive and relational skills, duration of illness (the time between the first psychiatric treatment and start of IMR program). Patient’s attendance to the group could be suspended if they found a job or returned to school with a schedule that was not compatible with IMR, as the priority was given to programs aimed towards their functional recovery. A further reason for excluding a patient from the group was a relapse in psychiatric symptoms or admission to the inpatient unit. Hence, the exact group size varied per group and per sessions.

The intervention was delivered in person by the provider. In some instance IMR was delivered individually due to either the user’s work schedule conflicting with the group calendar or to accommodate explicit user’s request.

During the first SARS-CoV-2 pandemic lockdown (March 2020–June 2020) IMR was delivered by remote (Starace and Ferrara 2020). During the pandemic the intervention was delivered individually to those patients who were attending IMR groups before the pandemic hit.

The groups were conducted by different professionals that attended the training, including psychiatrists, doctors in toxicology, psychologists, psychiatric rehabilitation technicians, nurses, social workers, professional educators. The groups were conducted at the local outpatient service that was providing care to the participating users.

Each session had a duration of 45–60 min and was delivered on a weekly basis. Breaks during the sessions were also an option because some attendee can have disturbing symptoms or limited attention span. The IMR intervention comprised eleven modules: M1. Recovery Strategies, M2. Basic Facts About Mental Illness M3. The Stress-Vulnerability Model, M4. Building social support, M5. Using Medication Effectively, M6. Drug and Alcohol Use, M7. Reducing Relapse, M8. Coping with Stress, M9. Coping with Persistent Symptoms, M10. Getting Your Needs Met in the Behavioral Health System M11. Healthy Lifestyles. All the didactic material, including the handouts, were available in Italian (Boggian et al. 2016). All the sessions (type of the module administered, duration) were adapted to the characteristics of the group and the service setting that was providing the intervention. For instance, the groups conducted at the SerD were predominantly focused on substance use and relapse prevention, while at EDP the module on substance abuse was not delivered.

The goal for each participant was to offer a group intervention that would allow for: (1) de-stigmatization and normalization of psychiatric illness, by acknowledging that other people share a similar personal history, (2) peer-support to overcome shared challenges during the process of recovery. Finally, the overall goal of the intervention was to increase illness awareness and improve personal recovery by providing tools to manage the disease.

Outcome Measures

Two outcome measures were adopted by the IMR providers: the IMR Scale (IMRS) and the IMR summary form.

The IMRS was administered at the end of the intervention as a further tool to foster a constructive discussion on personal recovery goals between user and provider.

This scale was developed to measure the effectiveness of the IMR intervention, and especially the link between illness management and recovery, in addition to the evaluation of the two dimensions taken individually (Hasson-Ohayon et al. 2008). There is a version of this scale intended for completion by the user and one intended for completion by the reference operator (Salyers et al. 2007). The scale consists of 15 items, scored on a Likert scale (Fardig et al. 2011) from 1 to 5. The outcome is expressed as the sum of the scores of the various items, thus ranging from 15 to 75. The higher the total score, the better the self-management of the disease (Salyers et al. 2007). Some studies also report the result as the average of the individual item scores, thus ranging from 1 to 5. The IMRS takes into account both the user's and the operator's perspective (Salyers et al. 2007) that seem to capture different but complementary perspectives of IMR themes (Fardig et al. 2011).

At the end of the intervention, the providers, in addition to their IMRS, also completed a structured IMR summary form: this was an online google module form, ideated by one of the local administrators, compiled by the provider at the end of each patient’s IMR program. The IMR summary form included information regarding the intervention, specifically the start date and stop date of the IMR program, the service where the program was delivered (addiction service, adult community mental health service, eating disorder service), IMR format (individual vs group), version of the manual adopted, number of sessions delivered, IMR modules administered, post-intervention user-IMRS and provider-IMRS scores.

Data collection

Anonymized socio demographic data (birth year, male/female birth sex, place of residence) and clinical information (date of first access to the mental health services, ICD-10 diagnosis, current prescribed medications, start date of current therapeutic treatment) were retrieved from the electronic health records (EHR) EFESO and Sister, respectively EHR of the Adult community mental health centers and Addiction services. Years of education, marital status, living condition, and employment status were recorded only at the first access to the psychiatric service. Intervention data were retrieved from the IMR summary form.

As this project was classified as an audit of current practice and service evaluation, full ethical approval was not required. However, it had been discussed with the local multi-disciplinary team, and management approval was obtained locally.

Statistical Analysis

A descriptive analysis of the main socio-demographic characteristics of the participants and the services that provided the intervention was conducted. Chi-square analyses were used for categorical variables (sex, place of residence, and psychiatric diagnosis), while t-tests, Mann Whitney U tests and analyses of variance (ANOVAs) were used for continuous variables, specifically the distribution of illness duration between services, the number of modules administered, and sessions conducted. Correlations between IMRS scores of users and providers were explored by Pearson's rank. Analyses were two-tailed with the significance level set to 0.5. All the analyses were performed in IBM SPSS Statistic.


Demographic Details

As detailed in Table 1, 126 patients completed the IMR intervention during the observation period. The majority were men (77; 61.1%), with a mean age of 40.14 years old (SD = 12.92), and five patients being 18 years old at the start of the IMR intervention. The mean age at the first mental health treatment was 34.55 years (SD = 12.06).

Table 1 Demographic and clinical characteristics of the IMR participants (N = 126)
Fig. 1
figure 1

Modules received by individual patients (N = 126), and Service that administered IMR. Each bar corresponds to a single patient. Each color represents a single module. On the x-axis the services that provided IMR is indicated, on the y-axis the number of the modules administered to each individual is reported. Abbreviations: CMHC (Community Mental Health Centers); EDP (Program for Eating Disorders); SerD (Service for addictions and Dependence).

Illness Details

Diagnoses varied greatly between services: as expected, patients diagnosed with substance use disorders were mostly attending IMR sessions at SerD, except for one patient receiving IMR at the CMHC. The most frequent psychiatric diagnosis was substance use disorders (42.1%). Duration of illness varied significantly between the services (p < 0.001). Specifically, the range was between < one year (for EDP) and 26 years. The most frequently prescribed drug treatment included antipsychotics (42.9%), and anxiolytics and sedatives (40.5%); of note, 20.6% of the sample was not receiving any psychotropic medication at the time of the IMR intervention.

Intervention Details

The IMR intervention was implemented in ten different services. Specifically, within the 126 patients, 69 (54.8%) were receiving IMR at one of the five CMHCs, 52 (42.1%) at the SerD, and 5 (4%) at EDP. The total duration of the implemented interventions was, on average, 9.6 months (SD = 7.5, [2–38]). Each patient received on average 21.4 sessions (SD = 12.6), however, the number of sessions varied considerably among services with a minimum of a single session/user delivered by a SerD to a maximum of 70 sessions in a single CHMC user (Table 2). Details on the implementation of IMR in the Ferrara DAISM-DP are summarized in Tables 2, 3.

Table 2 Services where IMR was implemented, and IMR Modules and sessions administered
Table 3 IMR scale (IMRS) at post-intervention: users and providers scores

As represented in Fig.1, the module of the IMR program that had been administered more frequently was the Module #1 “Recovery Strategies” (70.6% of patients received module 1), while the two that were administered the least were Module #9 “Coping with Persistent Symptoms” (6.3%) and Module #10 “Getting Your Needs Met in the Behavioral Health System” (6.3%). Moreover, module #6, which is related to drug and alcohol use, was the one most frequently used at SerD (administered to 100% of the SerD users), while it was rarely administered to the users of CMHCs (two patients received it). Interventions provided to patients with Eating Disorders diagnosis involved the adoption of modules #3 and #8 only, stress-vulnerability model and coping with stress module respectively.

The most frequently implemented modules included those related to drug and alcohol use (39.7%) and healthy lifestyles (43.7%), as well as those more general such as those related to recovery, knowledge of the disease, prevention of stress and relapse, and medication management. In summary, none of the services administered all modules, and no users received the entire IMR program (11 modules).

Outcome Details

As detailed in Table 3, at post-intervention the mean IMRS score (SD) of the individual items was 3.7 (0.5) and 3.6 (0.6) for the user and operator versions, respectively. The average total IMRS score was 55.1 (7.47) and 54.18 (8.45) for the user and operator versions, respectively. There was a strong positive statistically significant correlation between the operator and user scores (r = 0.96, n = 15, p = 0.001).


The aim of this study was to describe the implementation of the IMR Program within an entire department and to describe the outcomes achieved in terms of implementation process and users’ personal recovery.

Over a span of four years, the IMR intervention has been successfully completed by 126 patients with heterogeneous characteristics in terms of age, diagnosis, and setting. These preliminary findings show the feasibility of the exposure to IMR intervention in a real-world setting within the different sections of the Department, i.e., CMHCs, SerD, and EDP.

Contrary to what was originally established by the IMR Authors (Mueser et al. 2006), none of the services administered all 11 modules, thus none of the patients received all IMR modules. The module that appeared to have been administered the most was that dedicated to recovery strategies. Moreover, a trend towards preferences by certain services was observed: the module most frequently delivered in SerD was the module on drugs and alcohol, while it was rarely administered to the CMHC users. This means that conductors tailored their intervention based on the characteristics of the groups and the setting within which they were operating. This finding is in line with the theoretical framework and the programmatic plan overseen by the DAISM-DP with the indication to adapt the intervention to the different settings, and to tailor it around individuals and groups’ needs. It also responds to resource allocation constraints and time availability by certain participants. These challenges include staff turnover, demands on workforce competencies, competing role demands, and a lack of accessible and practical resources, highlighted in similar implementation studies (Mancini et al. 2009; Woltmann et al. 2008). Interventions delivered at CMHCs rarely included the module related to drug and alcohol use: this detail could reflect a shared choice by users and providers, as the topic was probably not of interest to the group. However, it is worth mentioning that, given the remarkable lifetime comorbidity of alcohol and substance use among patients with SMI (Murthy et al. 2019), it would be appropriate to integrate this module in CHMC and EDP as well.

Regarding IMR outcomes, the mean IMRS scores post-intervention showed a high correlation between those of the users and the operator meaning that the assessment on personal recovery made by users was in line with that made by the provider. This result is in line with that found in previous studies (Hasson-Ohayon et al. 2008; Salyers et al. 2007): these had indeed shown the presence of a positive, but low correlation consistent with the multidimensional nature of recovery (Salyers et al. 2007), which highlighted, according to the authors, the need to use both versions of the scale to fully capture the effectiveness of the IMR intervention by considering both points of view (Hasson-Ohayon et al. 2008; Salyers et al. 2007). Moreover, the post-intervention IMRS scores in our study were comparable to those provided by the published studies (user 3.7 (0.5) vs 3.6 (0.5); operator 3.6 (0.6) vs 3.5 (0.6).

It is difficult to compare our findings with other studies, as there is no evidence of the systematic implementation of IMR within an entire department of mental health, especially in Italy. One quasi-experimental study evaluated the implementation of IMR in different mental health services of the same region, namely six CMHC in Indiana, US (Salyers et al. 2009). They found that IMR could be implemented with a high degree of fidelity and in a short period of time. A more recent study, conducted in an inpatient unit of a single hospital showed that the addition of IMR to standard treatment improved personal recovery: although the sample was small (ten inpatients), this study highlighted the feasibility to implement IMR in acute settings (Miyajima et al. 2023).


This study establishes the feasibility of implementing IMR within three different community outpatient services of a single Department of Mental Health and Substance Abuse, with the intent to offer a therapeutic framework to mental health professionals regardless of their specific training or role within the department. The study shows that many professionals attended the IMR training and adopted it in their routine work, offering IMR to patients with different characteristics. Moreover, IMR was for the first time implemented in a service dedicated to eating disorders, and to individuals < 18 years old. Evidence about the implementation of IMR in children and adolescents service is missing, however our study shows that some modules, especially those dedicated to The Stress-Vulnerability Model and Coping with stress are deliverable to minors.

This is the first evidence of the effective implementation of IMR in Italian outpatient community mental health and substance abuse services, thus can represent a positive antecedent for other services to replicate the experience. The current study shows that an IMR-oriented approach could possibly be adapted to different users and services and could be further improved in terms of implementation and effectiveness.


The findings of the study should be interpreted in light of several limitations. First, this study is not randomized, and a control site was not present. Second, a possible selection bias could have happened, as the intervention might have been offered to those patients who were in a stable phase of the psychiatric illness, more motivated, with more time availability, thus those whose case manager thought they might complete the program successfully. Third, there is wide heterogeneity of sample size, setting, diagnosis, and sessions delivered, thus it is difficult to draw any conclusion about possible outcome differences between diagnostic categories or setting. One more limitation is that the information on how many patients IMR was offered to and did not participate, or those who dropped out was not available. A fidelity measure was also not implemented: this decision might have influenced the ample variability that we observed in the number of types of modules that was used during IMR; also, since the level of fidelity could have influenced the outcomes (Roosenschoon et al. 2021), future studies should consider to assess it systematically. Moreover a General Organizational Index scale that evaluates both the individualization of the program, and the quality improvement was not administered (McHugo et al. 2007).

The IMRS was the only structured clinical outcome measure adopted, and this limits the investigation on the clinical outcome of the intervention. Future studies should consider measuring clinical outcomes with standardized methods, such as the frequency and the length of hospitalizations, the symptoms severity, the global functioning, specific dimensions of recovery, and the caregiver’s burden.

This implementation did not include the administration of IMRS before the start of the intervention, but only after. Therefore, it is not possible to conduct a direct type of evaluation, that is, comparing IMRS scores at baseline with post-intervention scores, of program outcomes.

Finally, the findings presented in this study are limited to a specific catchment area and a single Department, thus finding might not be immediately comparable with different geographical or socio-economic areas.

Future Directions

Future studies should focus on the measurement of outcomes differentiated by psychiatric diagnosis, duration, and phase of illness, with specific attention to those diagnostic categories for whom IMR hasn’t been implemented so far (i.e., eating disorders, minors). A systematic analysis of the individual and setting predictors of outcomes would help to advance the field and would allow the mental health services to better tailor the intervention. For example, female patients might need tailored additional interventions (Ferrara and Srihari 2021), as the epidemiology and clinical characteristics of severe mental disorders differ considerably between the two sexes (Ferrara and Srihari 2021; Pallier et al. 2022; Pence et al. 2022).


This study shows the successful implementation of IMR intervention in a real-world setting. IMR was implemented in different outpatient mental health settings, with a great variability in terms of patients’ characteristics, and IMR module received. Future research should include a standardized measurement of clinical outcomes and identify possible outcome predictors to better tailor the intervention.