Introduction
Hoarding disorder (HD) is a mental health disorder characterised by difficulty discarding possessions, resulting in deterioration in living conditions, clutter, distress, and impairment to the individual’s functioning (American Psychiatric Association, 2013). HD often presents in adolescence (Tolin et al., Reference Tolin, Meunier, Frost and Steketee2010), with severity increasing with age (Ayers et al., Reference Ayers, Saxena, Golshan and Wetherell2010). Hoarding is linked to an increased risk of fire, increased damage when fires occur (London Fire Brigade, 2020), and a disproportionate number of fire-related fatalities (Chief Fire Officers Association, 2014; Lucini et al., Reference Lucini, Monk and Szlatenyi2009). There is evidence of an association between hoarding and an increased risk of accidental death (Waters et al., Reference Waters, Eckhardt and Eason2022). Hoarding has a significant impact on social care services, public health and systems around the individual with HD (Tolin et al., Reference Tolin, Frost, Steketee, Gray and Fitch2008; Bratiotis and Woody, Reference Bratiotis and Woody2020).
Individuals with HD are proposed to have vulnerabilities such as challenging early life experiences, genetic factors, and family influence. The individual is proposed to experience cognitive and information processing difficulties in relation to their decision making, attention, problem solving capacity and memory (Grisham et al., Reference Grisham, Norberg, Williams, Certoma and Kadib2010). These vulnerabilities and cognitive factors influence the beliefs that an individual develops in relation to objects and themselves, including their identity, values, responsibility, memory and control. These factors contribute to the saving and acquiring cycle, where positive (e.g. positive emotional responses to objects, or desire to acquire) or negative emotional responses (e.g. avoidance), reinforce of saving and acquiring (Steketee and Frost, Reference Steketee and Frost2007).
Cognitive behavioural therapy (CBT) is an empirical psychological intervention for HD (Bratiotis et al., Reference Bratiotis, Muroff and Lin2021; Frost and Hartl, Reference Frost and Hartl1996), with modifications including group and peer supported delivery (Bratiotis and Steketee, Reference Bratiotis and Steketee2015). A meta-analysis of CBT interventions (n=12) for HD found that whilst CBT typically leads to reliable change with a large effect size (g=0.82), it does not produce clinically significant change for the majority (57–76%) of participants (Tolin et al., Reference Tolin, Frost, Steketee and Muroff2015), although a more recent meta-analysis (Rodgers et al., Reference Rodgers, McDonald and Wootton2021) found a larger effect size (g=1.25) than Tolin and colleagues. A systematic review of interventions for HD with a broader scope [i.e. including psychopharmacology (n=3), psychological (CBT n=12), cognitive remediation (n=2), family interventions (Family Therapy, n=2) and online support groups (n=1)] reported significant improvements in hoarding symptoms across standardised outcome measures, for all included interventions (Thompson et al., Reference Thompson, de la Cruz, Mataix-Cols and Onwumere2017). However, this change was not clinically meaningful for most participants, who remained significantly impaired, scoring above clinical cut-offs for HD.
Few individuals with HD seek help from mental health services, with one study identifying that only 16% of their sample sought help (Robertson et al., Reference Robertson, Paparo and Wootton2020). Lack of knowledge about treatment (42%), its potential helpfulness (26%; Robertson et al., Reference Robertson, Paparo and Wootton2020) and acceptability (Rodriguez et al., Reference Rodriguez, Levinson, Patel, Rottier, Zwerling, Essock, Shuer, Frost and Simpson2016) of psychological interventions are cited as barriers to help seeking and engagement. Approximately 58% of individuals felt that they would prefer to resolve their hoarding difficulties without psychological support, while 28% of individuals with HD believed their hoarding difficulties did not require psychological intervention (Robertson et al., Reference Robertson, Paparo and Wootton2020). Help is often only sought in later life when difficulties have become more severe or support systems have been lost (Eckfield and Wallhagen, Reference Eckfield and Wallhagen2013; Mackin et al., Reference Mackin, Arean, Delucchi and Mathews2011).
Barriers to help seeking and engagement, in combination with limitations in achieving clinically meaningful change, raise questions regarding the utility of psychological interventions as the primary intervention for HD. Some individuals with HD may not be ready to access and use psychological interventions until their physical and safety needs have been met. Rodriguez et al. (Reference Rodriguez, Herman, Alcon, Chen, Tannen, Essock and Simpson2012) found that of people at risk of eviction due to severe HD, only half were receiving support for their mental health. Recent research has highlighted the potential importance of integrating harm reduction approaches to improve the safety of an individual before psychological intervention (David et al., Reference David, Crone and Norberg2021).
Researchers have begun to investigate the utility of interventions for HD with a psychosocial focus (Davidson et al., Reference Davidson, Dozier, Pittman, Mayes, Blanco, Gault, Schwarz and Ayers2019). Psychosocial interventions can be conceptualised by the inter-disciplinary biopsychosocial model, which highlights the importance of recognising the interaction between individuals, their physical and social context, and community is key (Egan et al., Reference Egan, Tannahill, Petticrew and Thomas2008; England et al., Reference England, Butler and Gonzalez2015; Martikainen et al., Reference Martikainen, Bartley and Lahelma2002). Psychosocial interventions can therefore aim to impact social, behavioural and environmental stressors, alongside promoting the development of adaptive coping strategies. A range of psychosocial interventions for HD have been developed and are delivered by non-healthcare professionals via local government organisations, including fire departments and social care services. Typically, these interventions have a practical focus aiming to reduce fire and health risk, increase housing stability and improve quality of life for the individual with HD. Such interventions generally have secondary benefits to the state of the property and the consequent impact on the community (Bratiotis et al., Reference Bratiotis, Woody and Lauster2019; Bratiotis and Woody, Reference Bratiotis and Woody2020).
To date, no reviews have synthesised the psychosocial interventions available for HD. In previous reviews of interventions for HD, psychosocial interventions related to fire safety and housing stability, delivered by allied health professionals (e.g. nurses, social workers and multi-disciplinary teams) have been excluded. Consequently, the relative effectiveness of these interventions is unknown. This is important given the reported reluctance from people with HD to seek support from mental health services and the likelihood that any support will instead be provided by other professionals. Thus, the aim of this review is to identify and synthesise what psychosocial approaches are available for HD, how and who delivers such interventions, and how effective these interventions are.
Method
The protocol for this systematic review was pre-registered on PROSPERO (https://www.crd.york.ac.uk/prospero/; CRD42021239453). The reporting of this review follows the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA; Page et al., Reference Page, McKenzie, Bossuyt, Boutron, Hoffman, Mulrow, Shamseer, Tetzladd, Akl, Brennan, Chou, Glanville, Grimshaw, Hróbjartsson, Lalu, Li, Loder, Mayo-Wilson, McDonald, McGuinness, Stewart, Thomas, Tricco, Welch, Whiting and Moher2021) and Synthesis Without Meta-analysis guidelines (SWiM; Campbell et al., Reference Campbell, McKenzie, Sowden, Katikireddi, Brennan, Ellis, Hartmann-Boyce, Ryan, Shepperd, Thomas, Welch and Thomson2020).
Inclusion and exclusion criteria
Participants
Inclusion criteria. Adults aged 18+ who meet HD diagnostic criteria or have been diagnosed with HD based upon DSM (American Psychiatric Association, 2013), ICD (World Health Organisation, 1993; World Health Organisation, 2019) criteria or who had scored above clinical cut-offs on standardised HD measures, e.g. Hoarding Rating Scale (HRS; Tolin et al., Reference Tolin, Meunier, Frost and Steketee2010); Savings Inventory-Revise (SI-R; Frost et al., Reference Frost, Steketee and Grisham2004); Clutter Image Rating (CIR; Frost et al., Reference Frost, Steketee, Tolin and Renaud2008), or whose homes were identified as premises impacted by hoarding.
Exclusion criteria. Participants were excluded if animal hoarding was identified as the primary HD difficulty. There were no exclusions based on participant co-morbidities, location, gender, sex, ethnicity, or medication usage, provided that HD was the primary mental health difficulty.
Intervention
Inclusion criteria. ‘Psychosocial’ was defined as: ‘pertaining to the influence of social factors on an individual’s mind or behaviour, and to the interrelation of behavioural and social factors’ (Oxford English Dictionary, 2012). Therefore, ‘Psychosocial Interventions’ will be included if they aim to change or influence the individual’s behavioural and social factors, with the primary outcome being psychosocial in nature (i.e. increasing housing stability, reducing fire risk, improving quality of life).
Exclusion criteria. Interventions primarily focused on animal hoarding were excluded. Psychological treatment interventions [e.g. CBT, Family Therapy, compassion focused therapy (CFT)] that were defined by the study authors as being such (e.g. this intervention is CBT/CFT) or where the modality of the intervention was not specifically named, but described components that clearly represented the type of therapy (e.g. an intervention must include change in cognitions that goes beyond learning practical skills to be considered CBT) were excluded.
Comparator/control
Inclusion criteria. All interventions were eligible as comparator/control conditions. However, not all included studies need to consist of a comparator/control intervention.
Outcome
Primary outcomes. The primary outcomes were psychosocial (i.e. focused on the psychosocial world of the individual with HD, i.e. quality of life, fire safety, housing stability, clutter in the home). This could be investigated using standardised outcome measures (e.g. Clutter Image Rating (CIR); Frost et al., Reference Frost, Steketee, Tolin and Renaud2008) or unstandardised measures (e.g. case outcome, fire code violations).
Secondary outcomes. Secondary outcomes could include standardised measures linked to hoarding symptomatology (e.g. Hoarding Rating Scale (HRS); Tolin et al., Reference Tolin, Meunier, Frost and Steketee2010).
Study design
Inclusion criteria. All study designs were included.
Identification of studies
Search strategy
Searches were conducted using three databases: PsycInfo (APA PsycNET); PubMed (MEDLINE); and Embase (Embase.com). Grey literature was searched using PsyArXiv. The first empirically based conceptual framework for HD was published in 1993 (Frost and Gross, Reference Frost and Gross1993), thus the range for the searchers was January 1993 to April 2022 (searches updated in January 2024, with no new records meeting inclusion criteria). The search strategy was designed to capture (1) psychosocial interventions for (2) adults with (3) HD using relevant keywords (see Supplementary material for full search strategy).
Selection of studies
Identified studies were screened for duplicates through Covidence Systematic Review Software (www.covidence.org), with additional duplicates manually identified and excluded. Eighty-five per cent of titles and abstracts were double-screened against the inclusion and exclusion criteria by two independent reviewers (D.T. and E.B.). Studies that passed the initial screen were reviewed as full text by two independent reviewers (D.T. and E.B.). The reference lists of all included studies were independently searched for relevant studies. Any discrepancies were discussed by the two reviewers, and consensus was reached. Where additional discussion was required, a third reviewer (J.M.) was consulted.
Data extraction
Two reviewers (D.T. and E.B.) independently extracted relevant data from all included studies. Data extracted include: title, setting, study design, study duration, inclusion and exclusion criteria, mean age and diagnosis (see Supplementary material for details of the data extraction criteria).
Quality assessment
Two reviewers (D.T. and E.B.) independently assessed the quality of all included studies using the Mixed Methods Appraisal Tool (MMAT; Hong et al., Reference Hong, Pluye, Fàbregues, Bartlett, Boardman, Cargo, Dagenais, Gagnon, Griffiths, Nicolau, O’Cathain, Rousseau and Vedel2018). The MMAT was chosen as it enables quality appraisal of all study designs. Included studies were first reviewed based on two screening questions, to determine if they met the criteria for empirical studies. This was followed by quality appraisal of the relevant methodological category. Any discrepancies were discussed to reach consensus. Quality assessment scores are presented as a fraction (Table 1), to illustrate the number of quality criteria assessed and met.
Table 1. Characteristics of included studies
Results
Searches
Searches of all sources retrieved n=365 records; of these, 43 were full text reviewed, resulting in five papers being included for review. The outcomes of all searches are reported in Fig. 1.
Figure 1. PRISMA 2020 diagram of study identification and selection.
Characteristics of included studies
Table 1 presents an overview of the characteristics of included studies.
Standardised hoarding disorder outcome measures
As identified within this systematic review of the literature, there are a range of standardised outcome measures that are used in the included studies, which are described further below. The SI-R (Frost et al., Reference Frost, Steketee and Grisham2004) is a 23-item questionnaire. This measure includes three subscales: clutter, difficulty discarding, and excessive acquisition. This measure is often used to inform diagnostic classification in HD. The clinical cut-off on this measure is 41 for the total score, with scores at or above this being suggestive of HD.
The Clutter Image Rating CIR (Frost et al., Reference Frost, Steketee, Tolin and Renaud2008) is an image-based scale which is used to aid assessment of clutter severity. Three residential rooms are the focus of this measure – living rooms, kitchens, and bedrooms. The CIR consists of nine numbered images for each room, with an increase in number correlating with an increase in clutter. The CIR can be used as self-report, or used by an external observer. On the CIR, a score of 4 or higher suggests an individual may need support for clutter linked to their HD.
The Hoarding Rating Scale (HRS; Tolin et al., Reference Tolin, Meunier, Frost and Steketee2010) is a 5-item questionnaire, which assesses clutter, difficulty discarding, excessive acquisition and the consequent distress and impairment. A clinical cut-off score of 14 has been recommended in the literature (Tolin et al., Reference Tolin, Meunier, Frost and Steketee2010), and is commonly utilised within research as a criterion for meeting HD diagnostic criteria.
The Health Obstacles Mental health Endangerment Structure and safety Risk Assessment Tool (HOMES; Bratiotis et al., Reference Bratiotis, Sorrentino Schmalisch and Steketee2011) allows for structured assessment of risk in hoarded homes. The tool considers risks of the impact of multiple psychosocial factors including safety of the home, risks to health and wellbeing of the person with HD and their family, as well as mental health factors.
Across the studies the most commonly used standardised outcome measures were the Saving Inventory-Revised (SI-R; n=1; Frost et al., Reference Frost, Steketee and Grisham2004), the Clutter Image Rating (CIR; n=4; Frost et al., Reference Frost, Steketee, Tolin and Renaud2008) and the Hoarding Rating Scale (HRS; n=1; Tolin et al., Reference Tolin, Meunier, Frost and Steketee2010). The HOMES (n=1; Bratiotis et al., Reference Bratiotis, Sorrentino Schmalisch and Steketee2011) was used as an assessment measure in one study, and was not repeated at discharge (Metropolitan Boston Housing Partnership, 2015).
Participants
HD diagnosis was confirmed using the Structured Interview for Hoarding Disorder (n=2; Millen et al., Reference Millen, Levinson, Linkovski, Sauer, Thaler, Nick, Johns, Vargas, Rottier, Joyner, Gibson, Zierling, Sonnenfeld, Shapiro, Tannen, Conover, Essock, Herman, Simpson and Rodriguez2020; Pittman et al., Reference Pittman, Davidson, Dozier, Blanco, Baer, Twamley, Mayes, Sommerfeld, Lagare and Ayers2021), or by professional inspection of the participant’s home residence, which may include measures such as the CIR (n=3; Kwok et al., Reference Kwok, Bratiotis, Luu, Lauster, Kysow and Woody2018; Kysow et al., Reference Kysow, Bratiotis, Lauster and Woody2020; Metropolitan Boston Housing Partnership, 2015). Only one study specified that their aim was to recruit a sample of older adults (age 60 years; Pittman et al., Reference Pittman, Davidson, Dozier, Blanco, Baer, Twamley, Mayes, Sommerfeld, Lagare and Ayers2021). However, across studies, the mean age was 62.46 years, categorising participants on average as older adults. Across the five studies, a majority of participants were female, with only one study having a majority male sample (Kwok et al., Reference Kwok, Bratiotis, Luu, Lauster, Kysow and Woody2018). A majority of participants were renting in three studies, but only the Metropolitan Boston Housing Partnership (2015) study focused intentionally on individuals who were renting their homes. Two studies did not state housing status (Millen et al., Reference Millen, Levinson, Linkovski, Sauer, Thaler, Nick, Johns, Vargas, Rottier, Joyner, Gibson, Zierling, Sonnenfeld, Shapiro, Tannen, Conover, Essock, Herman, Simpson and Rodriguez2020; Pittman et al., Reference Pittman, Davidson, Dozier, Blanco, Baer, Twamley, Mayes, Sommerfeld, Lagare and Ayers2021).
Study design and setting
No studies employed a randomised control design. Instead, all studies utilised a full or partial quantitative design, with three studies consisting of exploratory data analysis, and two pre–post quasi-experimental studies. The study which included qualitative methods (n=1; Pittman et al., Reference Pittman, Davidson, Dozier, Blanco, Baer, Twamley, Mayes, Sommerfeld, Lagare and Ayers2021), utilised mixed methods. All studies utilised community-based samples.
Quality appraisal of included studies
Table 1 presents the summary quality assessment score for included studies. All studies passed the two screening questions of the MMAT (Hong et al., Reference Hong, Pluye, Fàbregues, Bartlett, Boardman, Cargo, Dagenais, Gagnon, Griffiths, Nicolau, O’Cathain, Rousseau and Vedel2018). Overall, the quality of studies was good, with all studies meeting at least three out of five of the quality assessment criteria.
A total of four different interventions were identified across the included studies. The Hoarding Action Response Team (HART) model was utilised in two studies (Kwok et al., Reference Kwok, Bratiotis, Luu, Lauster, Kysow and Woody2018; Kysow et al., Reference Kysow, Bratiotis, Lauster and Woody2020). Other interventions included Hoarding Intervention and Tenancy Preservation Project (HI/TPP; Metropolitan Boston Housing Partnership, 2015), Critical Time Intervention-HD (CTI-HD; Millen et al., Reference Millen, Levinson, Linkovski, Sauer, Thaler, Nick, Johns, Vargas, Rottier, Joyner, Gibson, Zierling, Sonnenfeld, Shapiro, Tannen, Conover, Essock, Herman, Simpson and Rodriguez2020) and Cognitive Rehabilitation and Exposure/Sorting Therapy (CREST) Community Program (CCP; Pittman et al., Reference Pittman, Davidson, Dozier, Blanco, Baer, Twamley, Mayes, Sommerfeld, Lagare and Ayers2021). Table 2 presents an overview of the characteristics of the four included interventions.
Table 2. Characteristics of included interventions
Practitioners involved in the interventions
Three studies reported multi-disciplinary teams of professionals delivering the interventions (Kwok et al., Reference Kwok, Bratiotis, Luu, Lauster, Kysow and Woody2018, Kysow et al., Reference Kysow, Bratiotis, Lauster and Woody2020; Pittman et al., Reference Pittman, Davidson, Dozier, Blanco, Baer, Twamley, Mayes, Sommerfeld, Lagare and Ayers2021). Such teams consisted of psychologists and family therapists (Pittman et al., Reference Pittman, Davidson, Dozier, Blanco, Baer, Twamley, Mayes, Sommerfeld, Lagare and Ayers2021), registered psychiatric nurses (n=1; Kwok et al., Reference Kwok, Bratiotis, Luu, Lauster, Kysow and Woody2018), fire officers (n=2; Kwok et al., Reference Kwok, Bratiotis, Luu, Lauster, Kysow and Woody2018; Kysow et al., Reference Kysow, Bratiotis, Lauster and Woody2020), peer support workers (n=1; Pittman et al., Reference Pittman, Davidson, Dozier, Blanco, Baer, Twamley, Mayes, Sommerfeld, Lagare and Ayers2021), healthcare workers (n=1; Kwok et al., Reference Kwok, Bratiotis, Luu, Lauster, Kysow and Woody2018), social workers (n=1; Pittman et al., Reference Pittman, Davidson, Dozier, Blanco, Baer, Twamley, Mayes, Sommerfeld, Lagare and Ayers2021), clinical health supervisors (n=1; Kysow et al., Reference Kysow, Bratiotis, Lauster and Woody2020) and property inspectors (n=1; Kwok et al., Reference Kwok, Bratiotis, Luu, Lauster, Kysow and Woody2018). However, two studies (Metropolitan Boston Housing Partnership, 2015; Millen et al., Reference Millen, Levinson, Linkovski, Sauer, Thaler, Nick, Johns, Vargas, Rottier, Joyner, Gibson, Zierling, Sonnenfeld, Shapiro, Tannen, Conover, Essock, Herman, Simpson and Rodriguez2020) did not distinguish professional background, classifying practitioners as ‘case managers’.
Control conditions
None of the included studies utilised control groups or comparator interventions.
Effectiveness of the interventions
Effectiveness data have been separated based on the primary and any additional outcome measures used within each report. Table 3 presents available pre- and post-data for primary and secondary outcomes, with effect sizes presented where available. Attrition information is also presented where provided within the included studies.
Table 3. Effectiveness and attrition rates for included studies
Effect size data were not available or calculable for three of the five studies. Of the two studies where data were available, effect sizes were large (see Table 3). The interventions of both Kysow et al. (Reference Kysow, Bratiotis, Lauster and Woody2020) and Pittman et al. (Reference Pittman, Davidson, Dozier, Blanco, Baer, Twamley, Mayes, Sommerfeld, Lagare and Ayers2021) showed a large effect size for reduction in clutter, with Pittman et al. also showing a large impact on hoarding symptomology. Attrition was relatively consistent across the studies (20–22%), except for Kwok et al. (Reference Kwok, Bratiotis, Luu, Lauster, Kysow and Woody2018) where it was reported to be zero, likely due to the mandatory nature of fire legislation enforcement. The most effective intervention based on eviction rates was CTI-HD, where no participants were evicted (Millen et al., Reference Millen, Levinson, Linkovski, Sauer, Thaler, Nick, Johns, Vargas, Rottier, Joyner, Gibson, Zierling, Sonnenfeld, Shapiro, Tannen, Conover, Essock, Herman, Simpson and Rodriguez2020), although the sample was very small (n=14).
In all studies, the risk of homelessness was identified. However, this was only a primary outcome for three of the five studies (Kwok et al., Reference Kwok, Bratiotis, Luu, Lauster, Kysow and Woody2018; Kysow et al., Reference Kysow, Bratiotis, Lauster and Woody2020; Metropolitan Boston Housing Partnership, 2015), with one study considering this a secondary outcome (Millen et al., Reference Millen, Levinson, Linkovski, Sauer, Thaler, Nick, Johns, Vargas, Rottier, Joyner, Gibson, Zierling, Sonnenfeld, Shapiro, Tannen, Conover, Essock, Herman, Simpson and Rodriguez2020), and the other only collecting homelessness risk data at baseline (Pittman et al., Reference Pittman, Davidson, Dozier, Blanco, Baer, Twamley, Mayes, Sommerfeld, Lagare and Ayers2021). One study did not declare evictions (CCP; Pittman et al., Reference Pittman, Davidson, Dozier, Blanco, Baer, Twamley, Mayes, Sommerfeld, Lagare and Ayers2021), and one had no evictions within their sample (Millen et al., Reference Millen, Levinson, Linkovski, Sauer, Thaler, Nick, Johns, Vargas, Rottier, Joyner, Gibson, Zierling, Sonnenfeld, Shapiro, Tannen, Conover, Essock, Herman, Simpson and Rodriguez2020). Both of these quasi-experimental studies acknowledged risk of homelessness within their sample, with the CTI-HD intervention specifying that the risk was still present post-intervention. This suggests that the change may not have been clinically significant, as the mean score on the CIR (mean change from 6 to 5) had not reduced below the clinical cut-off of 4. This supports the authors’ assertion that despite there being evidence of positive impact upon eviction, and hoarding symptoms, alternative and more effective treatments are needed for individuals with HD. In contrast to this, the large samples from service evaluations of HART (n=503) and HI/TPP (n=175) show a broader view on eviction and homelessness despite intervention. The data suggest that HI/TPP is up to 98% successful in preventing eviction (Metropolitan Boston Housing Partnership, 2015), with the results from HART suggesting 90% (Kysow et al., Reference Kysow, Bratiotis, Lauster and Woody2020) to 87% (Kwok et al., Reference Kwok, Bratiotis, Luu, Lauster, Kysow and Woody2018) success. For those evictions that took place, 75% began at too late a stage for the intervention to make enough change to prevent or delay proceedings (Kysow et al., Reference Kysow, Bratiotis, Lauster and Woody2020), or the participants withdrew from the intervention (Metropolitan Boston Housing Partnership, 2015).
Discussion
The aim of this systematic review was to investigate and synthesise the available literature on psychosocial interventions for individuals with HD, how and who delivers such interventions, and how effective these interventions are. Five studies were included detailing four interventions. These interventions were delivered by teams of multi-disciplinary professionals (n=3) or case managers (n=2). Whilst none of the interventions included in this systematic review was designed based on the CBT model for HD, this conceptualisation underpins the current understanding of HD. We will therefore draw on the CBT model for HD to illustrate the areas of the model that psychosocial interventions target. Thus, when applying the model, the psychosocial interventions can be viewed as targeting three components of the CBT model for HD (Frost and Hartl, Reference Frost and Hartl1996; Steketee and Frost, Reference Steketee and Frost2007): vulnerabilities; cognitive difficulties; and saving and acquiring. However, a distinction in the focus should be noted, with psychosocial interventions impacting upon the consequences of hoarding rather than precipitating factors. It is therefore likely to be important for psychosocial and CBT interventions to be used in conjunction.
A previous meta-analysis of CBT for hoarding suggested a large effect size for total HD severity (Tolin et al., Reference Tolin, Frost, Steketee and Muroff2015), but acknowledged the lack of clinically significant change within the samples. A more recent meta-analysis (Rodgers et al., Reference Rodgers, McDonald and Wootton2021) found a larger mean effect size for CBT for HD (g=1.25) but did not consider whether the change for participants within these studies was clinically significant. The results from the present review suggest that psychosocial interventions can produce large effects sizes for positive change in levels of clutter (Kysow et al., Reference Kysow, Bratiotis, Lauster and Woody2020 and Pittman et al., Reference Pittman, Davidson, Dozier, Blanco, Baer, Twamley, Mayes, Sommerfeld, Lagare and Ayers2021) and hoarding symptomology (Pittman et al., Reference Pittman, Davidson, Dozier, Blanco, Baer, Twamley, Mayes, Sommerfeld, Lagare and Ayers2021) that are comparable to similar outcomes from CBT interventions. This is promising, as it suggests that psychosocial interventions can produce similar change for participants to CBT interventions in addition to affecting change in other outcome areas such as eviction rates. Reducing housing evictions is a key outcome within most of the included studies in this review. The results show the importance of multi-disciplinary interventions for individuals with HD, as through specialised support, individuals could improve their safety and their living conditions to a level where they could withdraw or be discharged from the service. The HART (Kysow et al., Reference Kysow, Bratiotis, Lauster and Woody2020) and HI/TPP (Metropolitan Boston Housing Partnership, 2015) interventions provide further evidence for the impact of psychosocial intervention on the home; scores on the Clutter Image Ratings Scale (Frost et al., Reference Frost, Steketee, Tolin and Renaud2008) reduced from the clinical to the non-clinical range in both studies. Psychosocial interventions appear to produce observable change in the participants’ environments, reducing risk to safety and improving access to cooking and washing facilities. In line with Maslow’s hierarchy of needs (Maslow, Reference Maslow1943), this may then enable individuals to engage more successfully with psychological interventions such as CBT following a psychosocial intervention.
In studies of CBT for HD, attrition rates are variable. Some studies have reported attrition of approximately 33% (Gillam et al., Reference Gillam, Norberg, Villavicencio, Morrison, Hannan and Tolin2011), with others declaring no participant withdrawal (Ayers et al., Reference Ayers, Wetherell, Golshan and Saxena2011). A recent randomised trial found attrition rates for the CBT condition of 26% (Tolin et al., Reference Tolin, Wootton, Levy, Hallion, Worden, Diefenbach, Jaccard and Stevens2019). In this trial, 17% of participants were removed for non-compliance, whereas 9% made the choice to withdraw (Tolin et al., Reference Tolin, Wootton, Levy, Hallion, Worden, Diefenbach, Jaccard and Stevens2019). An investigation of attrition rates in HD (Ayers et al., Reference Ayers, Pittman, Davidson, Dozier, Mayes and Almklov2018) found that baseline clutter ratings, combined with denial of hoarding as a problem, predicted attrition. Attrition rates for the studies included in this review ranged from 0 to 22%. Across the studies, where reported, participant withdrawal occurred due to factors including potential pre-contemplative stage of change (limited acceptance of severity, consequences of HD, and harm to self; Metropolitan Boston Housing Partnership, 2015), motivation difficulties (Metropolitan Boston Housing Partnership, 2015; Pittman et al., Reference Pittman, Davidson, Dozier, Blanco, Baer, Twamley, Mayes, Sommerfeld, Lagare and Ayers2021), engagement difficulties (Kwok et al., Reference Kwok, Bratiotis, Luu, Lauster, Kysow and Woody2018; Kysow et al., Reference Kysow, Bratiotis, Lauster and Woody2020; Pittman et al., Reference Pittman, Davidson, Dozier, Blanco, Baer, Twamley, Mayes, Sommerfeld, Lagare and Ayers2021), health issues (Pittman et al., Reference Pittman, Davidson, Dozier, Blanco, Baer, Twamley, Mayes, Sommerfeld, Lagare and Ayers2021) and emotional impact (Metropolitan Boston Housing Partnership, 2015) of the intervention. Due to the motivational and pre-contemplative stage of change difficulties associated with withdrawal, it may be beneficial to consider augmenting the psychosocial interventions included in this review with motivational interviewing (Rollnick and Miller, Reference Rollnick and Miller1995) for those who are more ambivalent about change.
Studies generally presented demographic data for the samples, with comparisons to local population demographics where available (Kwok et al., Reference Kwok, Bratiotis, Luu, Lauster, Kysow and Woody2018; Kysow et al., Reference Kysow, Bratiotis, Lauster and Woody2020; Metropolitan Boston Housing Partnership, 2015), to consider whether they were meeting the needs of the local population, or if any demographic groups were missed. Such demographic observation led to expansion of the CCP intervention to include Spanish-speaking staff (Pittman et al., Reference Pittman, Davidson, Dozier, Blanco, Baer, Twamley, Mayes, Sommerfeld, Lagare and Ayers2021). However, there were some inherent limits for studies based on eligibility criteria or service level agreements such as participant economic status (Metropolitan Boston Housing Partnership, 2015; Pittman et al., Reference Pittman, Davidson, Dozier, Blanco, Baer, Twamley, Mayes, Sommerfeld, Lagare and Ayers2021) and severity of hoarding presentation (Kwok et al., Reference Kwok, Bratiotis, Luu, Lauster, Kysow and Woody2018; Kysow et al., Reference Kysow, Bratiotis, Lauster and Woody2020). Included participants may have been influenced by the referral methods for the studies, with older adults, for example, being more likely to be involved with social care services.
Participants within these samples were older than is typical for the populations in which they live, therefore age is an important factor to consider in relation to HD and symptom severity. In HD, there is typically an increase in severity of HD presentation as age increases (Ayers et al., Reference Ayers, Saxena, Golshan and Wetherell2010), potentially influenced by the available time for clutter to accumulate. Without treatment, HD presentations typically worsen over time, and be exacerbated by difficulties typically associated older adulthood, such as cognitive difficulties (Ayers et al., Reference Ayers, Dozier, Wetherell, Twamley and Schiehser2016) and loss of social support (Mackin et al., Reference Mackin, Arean, Delucchi and Mathews2011). Older adults with HD are more likely to have increased difficulties attributable to hoarding, such as an increase in clutter volume (Ayers et al., Reference Ayers, Najmi, Mayes and Dozier2015). It is therefore unsurprising that participants in the included studies were primarily adults from older age groups (60+). More significant difficulties are more likely to be identified and referrals made to specialist services, as supported by the help seeking in HD literature (Eckfield and Wallhagen, Reference Eckfield and Wallhagen2013; Mackin et al., Reference Mackin, Arean, Delucchi and Mathews2011). As both cognitive difficulties and loss of social support are implicated in ageing HD populations, it makes sense that psychosocial interventions may produce a significant impact. However, it would be useful to investigate whether individuals with reduced social support would self-select for psychosocial interventions if service coordination and access is promoted.
A limitation of many of the psychosocial interventions is that they are not based upon any particular theory. However, the outcomes are understandable when considered in the context of CBT conceptualisations of HD (e.g. Steketee and Frost, Reference Steketee and Frost2007). As depicted in Fig. 2, the psychosocial interventions identified enacted change in three areas of the CBT model for HD: vulnerabilities, cognitive processes and difficulties, and saving and acquiring. It is likely that these changes may also have influenced the emotional experiences, beliefs and meanings around hoarding, although these were not directly targeted by the interventions.
Figure 2. CBT model for hoarding, with additions indicating where psychosocial approaches may intervene and encourage change (adapted from Steketee and Frost, Reference Steketee and Frost2007).
In the CBT for hoarding model, vulnerabilities include early life experiences, genetic factors, familial and social influences. All included studies target the familial and social influences component of the vulnerabilities. Three of the five studies included aspects which coordinated or educated families and support networks (Kysow et al., Reference Kysow, Bratiotis, Lauster and Woody2020; Millen et al., Reference Millen, Levinson, Linkovski, Sauer, Thaler, Nick, Johns, Vargas, Rottier, Joyner, Gibson, Zierling, Sonnenfeld, Shapiro, Tannen, Conover, Essock, Herman, Simpson and Rodriguez2020; Pittman et al., Reference Pittman, Davidson, Dozier, Blanco, Baer, Twamley, Mayes, Sommerfeld, Lagare and Ayers2021). All of the interventions aimed to support the individual to access services, benefits and support (Kwok et al., Reference Kwok, Bratiotis, Luu, Lauster, Kysow and Woody2018; Kysow et al., Reference Kysow, Bratiotis, Lauster and Woody2020; Metropolitan Boston Housing Partnership, 2015; Millen et al., Reference Millen, Levinson, Linkovski, Sauer, Thaler, Nick, Johns, Vargas, Rottier, Joyner, Gibson, Zierling, Sonnenfeld, Shapiro, Tannen, Conover, Essock, Herman, Simpson and Rodriguez2020; Pittman et al., Reference Pittman, Davidson, Dozier, Blanco, Baer, Twamley, Mayes, Sommerfeld, Lagare and Ayers2021). All studies provided additional social interaction through involvement in the intervention and meetings with professionals.
Cognitive processes and difficulties associated with HD are incorporated within the CBT for hoarding model. This was not considered within all the included studies. The CCP (Pittman et al., Reference Pittman, Davidson, Dozier, Blanco, Baer, Twamley, Mayes, Sommerfeld, Lagare and Ayers2021) intervention included seven sessions of compensatory cognitive training, which included memory, planning, problem solving, and cognitive flexibility. HI/TPP specifically identified incorporated developing skills and strategies for organising (Metropolitan Boston Housing Partnership, 2015), with other interventions not specifically including cognitive strategies, unless included within optional self-help groups (Millen et al., Reference Millen, Levinson, Linkovski, Sauer, Thaler, Nick, Johns, Vargas, Rottier, Joyner, Gibson, Zierling, Sonnenfeld, Shapiro, Tannen, Conover, Essock, Herman, Simpson and Rodriguez2020).
Saving and acquiring are two of the most visible difficulties in HD, contributing to challenging levels of clutter. All interventions included components aimed to influence this aspect of the hoarding model. Exposure to discard is implicit within interventions for hoarding where an individual needs to reduce the impact of clutter on their lives. Within the included studies, two provided structured support for discard as standard (Metropolitan Boston Housing Partnership, 2015; Pittman et al., Reference Pittman, Davidson, Dozier, Blanco, Baer, Twamley, Mayes, Sommerfeld, Lagare and Ayers2021) or as an optional component of the intervention (Millen et al., Reference Millen, Levinson, Linkovski, Sauer, Thaler, Nick, Johns, Vargas, Rottier, Joyner, Gibson, Zierling, Sonnenfeld, Shapiro, Tannen, Conover, Essock, Herman, Simpson and Rodriguez2020). It was not specified within the HART interventions whether support for discard and clutter reduction was included; however, participants were able to request motivational and problem-solving support related to discard (Kysow et al., Reference Kysow, Bratiotis, Lauster and Woody2020), and clean-outs occurred independently for 28.9% of participants (Kysow et al., Reference Kysow, Bratiotis, Lauster and Woody2020).
Limitations
Due to the methodological heterogeneity (study design, components of intervention and outcomes assessed), and data reporting deficits, meta-analysis was not possible.
The quality of the studies within this review was generally good with all included studies meeting at least 3/5 quality assessment criteria. However, a particular deficit to note is the inconsistency in data reported, with a lack of applicable effect sizes presented, and some studies not presenting numerical data (Millen et al., Reference Millen, Levinson, Linkovski, Sauer, Thaler, Nick, Johns, Vargas, Rottier, Joyner, Gibson, Zierling, Sonnenfeld, Shapiro, Tannen, Conover, Essock, Herman, Simpson and Rodriguez2020), or not providing the data needed to calculate effect sizes (Kwok et al., Reference Kwok, Bratiotis, Luu, Lauster, Kysow and Woody2018; Metropolitan Boston Housing Partnership, 2015). Whilst these studies have illustrated the work that is being done by services to support individuals with HD, reporting effectiveness is vital to ensure comprehensive evaluation of interventions.
Implications for future research
Research in the field of HD is limited. Whilst it has begun to develop and expand, there is substantial room for further research that employs rigorous methodologies (Bratiotis et al., Reference Bratiotis, Muroff and Lin2021) Whilst location specific programmes like HI/TPP and HART have formalised and clarified their psychosocial intervention processes, HD is a global and cross-cultural problem (Fernández de la Cruz et al., Reference Fernández de la Cruz, Nordsletten and Mataix-Cols2016). There is opportunity for exploratory data analysis, or intervention development, in a multitude of regions and areas of expertise.
Additional and longitudinal service evaluations are needed. There has been limited data collected on long-term effectiveness of psychosocial interventions, with studies generally presenting follow-up periods with a maximum of 6 months in duration (Rodgers et al., Reference Rodgers, McDonald and Wootton2021). To consider effectiveness over time, longer follow up periods (e.g. 1–5 years post intervention) are required. Studies should consider whether any of the original service users re-present in subsequent analysis periods. Similarly, there is an absence of randomised control trials comparing psychosocial interventions against routine care and/or alternative interventions such as CBT.
Most participants identified within this review were older adults, with additional difficulties in relation to their psychosocial functioning and social support associated with age. It would therefore be pertinent to consider the implications of earlier intervention for individuals with HD and trial psychosocial interventions for adults of working age.
This review has highlighted the effectiveness of psychosocial interventions for people with HD. Future research, which considers both psychological and psychosocial interventions, and the order in which these are delivered, may help to produce coherent and integrative practitioner guidance.
Implications for policy and practice
It is important to consider the cost of HD to services, and how the integration of psychosocial interventions may impact this. Whilst analysis of cost aspects of psychosocial interventions was beyond the scope of this review, in the UK, data collected from local housing providers in the North-East region suggests a potential cost of more than £1.5 million over one year, for fewer than 150 identified individuals with HD (Neave et al., Reference Neave, Caiazza, Hamilton, McInnes, Saxton, Deary and Wood2017). Further costs in the region of £100,000 per year was attributed to HD by local fire and rescue services.
There is limited information regarding cost-effectiveness of the psychosocial interventions. However, the HI/TPP project presented a cost of around $1800 USD (Metropolitan Boston Housing Partnership, 2015) per client for longer term cases, and planned to expand the service. The CCP (Pittman et al., Reference Pittman, Davidson, Dozier, Blanco, Baer, Twamley, Mayes, Sommerfeld, Lagare and Ayers2021) and HART (Kysow et al., Reference Kysow, Bratiotis, Lauster and Woody2020) interventions were both extended beyond their original term.
When looking at services such as HART and HI/TPP, focused on tenancy preservation, there were high levels of success of avoiding eviction (98%; Metropolitan Boston Housing Partnership, 2015). For these studies, the reduction in other hoarding-related symptoms linking to the CBT model of hoarding was secondary to a reduction in clutter and improvement in living environment.
Conclusion
The efficacy of CBT for HD is modest (Tolin et al., Reference Tolin, Frost, Steketee and Muroff2015), therefore interventions that go beyond the standard CBT paradigm are important for CBT therapists to know about to guide practice, either in considering different approaches as alternative or adjunctive interventions to CBT. This systematic review indicates preliminary evidence to suggest that psychosocial interventions can support people with HD to enact change to prevent homelessness or eviction. However, substantial further research and evaluation is needed to aid the development of best practice guidelines for HD.
Key practice points
-
(1) There is evidence for the effectiveness of psychosocial interventions across a range of outcomes and beyond those typically measured in CBT intervention studies.
-
(2) When providing therapy for hoarding difficulties, CBT practitioners should consider working as part of a multi-disciplinary team.
-
(3) There are similarities in the targets and methods of psychosocial and CBT interventions for hoarding disorder. However, differences lie in how they are delivered and by whom.
-
(4) Further research is needed to develop the evidence base for psychosocial interventions for hoarding disorder and in particular there is a need for randomised control trials comparing this approach with routine care and/or active interventions, such as CBT.
Supplementary material
To view supplementary material for this article, please visit https://doi.org/10.1017/S1754470X24000357
Data availability statement
The authors confirm that the data supporting the findings of this systematic review are available within the article and its Supplementary material.
Acknowledgements
None.
Author contributions
Daisy Twigger: Conceptualization (equal), Data curation (lead), Formal analysis (lead), Investigation (lead), Methodology (equal), Project administration (lead), Validation (lead), Visualization (lead), Writing - original draft (lead), Writing - review & editing (lead); James Gregory: Conceptualization (supporting), Formal analysis (supporting), Investigation (supporting), Supervision (supporting), Writing - review & editing (supporting); Emma Bowers: Formal analysis (supporting), Writing - review & editing (supporting); Josie Millar: Conceptualization (equal), Formal analysis (supporting), Investigation (supporting), Methodology (equal), Project administration (supporting), Supervision (lead), Visualization (supporting), Writing - original draft (supporting), Writing - review & editing (supporting).
Financial support
This research received no specific grant from any funding agency, commercial or not-for-profit sectors.
Competing interests
The authors declare none.
Ethical standards
The authors have abided by the Ethical Principles of Psychologists and Code of Conduct as set out by the BABCP and BPS.
Open access
Comments
No Comments have been published for this article.