The Hourglass Model: Are There Structural Problems with the Scarcity of Positive Results for Flexible ACT?#

INTRODUCTION

A commentary was presented by van Veldhuizen, Delespaul and Mulder [1] regarding the critical review performed by Nordén and Norlander based on the five available empirical articles about Flexible Assertive Community Treatment (FACT) [2]. In the review we argued that the five studies could not show that FACT led to improvements for the clients in terms of symptoms, function, or wellbeing. The conclusion drawn was that at the present time there exists no evidence for FACT.

We view it as a positive thing that Veldhuizen et al. incorporated some of our views. For example, they agree that ”there is insufficient evidence for the effectiveness of Flexible ACT”, and that the Resource Group Assertive Community Treatment (RACT) method with its systematic ”involvement of the client’s personal network and external support services is a possible enrichment of Flexible ACT teams” [1]. However, at the same time the authors of the commentary avoid a discussion of the obvious lack of positive results despite extensive research during at least seven years [3-7] and for this reason an analysis of why FACT did not fare better is missing. This notion as well as the fact that an entirely new article from the Dutch group has been published [8], makes it necessary for us to return to the discussion with some comments. Unfortunately the new article exhibits a comparable lack of clarity as the previous five studies.

First, however, we would like to examine a few critical views which Veldhuizen et al. direct at us, views which we perceive as lacking in solidity. For example they argue that we ”disregard the significant positive findings concerning reduced service utilization by clients” [1] in the study by Firn et al. [7]. We cannot agree at all. In fact we argued that this British study showed several interesting results and we even cited the descriptions of the results by the authors themselves. In this case the problem was that it was unclear what the pre-and post- conditions actually measured i. e., it was not clear how much standard care, assertive outreach, or FACT was involved in the assessments. The British authors themselves suggested openly that the model they implemented ”is not equivalent to FACT teams in the Netherlands” [7]. Yet another objection from van Veldhuizen et al. was that the meta-analysis of RACT [9] had not examined the RACT model as a whole but only one or a few of its elements. This is not true. On the contrary, the meta-analysis which contains articles from the period 2001 – 2011, had strict inclusion criteria, one of which was that all the studies included were to be based on the manual of the model. An additional objection, difficult to understand, put forward by van Veldhuizen et al, to our critical review was that we supposedly ”ignored the study of the relationship between model fidelity and outcomes” [1] authored by van Vugt and associates. However, it is not clear in the article [10] whether any FACT teams were involved. Only ACT teams were mentioned. Furthermore, the DACT scale [11] was used in the assessments of ACT fidelity, as it should be!

THE SIXTH EMPIRICAL FACT ARTICLE

Our review [2] examined the five empirical FACT articles published at the time. Only recently a sixth empirical article has been published on FACT involving three teams and 372 clients. In the abstract the results are summarized as follows: ”Statistically significant improvements were found in compliance, unmet needs and quality of life…The percentage of remissions increased with 9 %" [8]. However, what should be seen as the main results did not yield a picture as positive as that. Although this time fully correct criteria for remission were used but still the result was not significant. In addition there was no significant effect for the well-established international instrument, Health of the Nation Outcome Scales (HoNoS) on either total score or the sub scales Symptoms, Behavior, Social problems or Impairments! Given that it turned out that three FACT teams of the study had good fidelity scores (in accordance with the original FACT fidelity scale) even at the outset and that two of the three teams had ”excellent” results at the completion of the study, it is therefore possible for us to study FACT teams which are functioning at their best. However, it was not possible to relate fidelity scores to outcomes. The conclusion of our review [2] is thus still valid, i. e., FACT cannot demonstrate progress in the treatment of the clients, in a persuasive way.

THE HOURGLASS MODEL

Even if van Veldhuizen et al. and the current authors now agree that there is insufficient evidence for the effectiveness of Flexible ACT, and that RACT may provide new impulses for FACT it is completely necessary to conduct a scientific analysis of why FACT, despite great national backing in the Netherlands and despite now extensive research, cannot provide better results. There is a large and exciting scientific question that needs to be addressed before we consider transforming FACT into some kind of RACT-FACT hybrid which apparently some researchers are now considering. It is important to analyze which components are positively and negatively useful, respectively, in a model. In the RACT model, such analyses of components are currently being carried out and one report has been published [12]. Given the relatively meagre results for FACT, it is imminent that such analyses are conducted even there. Here, we would like to raise the question whether there exists a structural problem with FACT.

In the FACT manual, authored by van Veldhuizen and Bähler [13], the main idea emerges that one single team strives to follow the patient through the entire chain of care and rehabilitation: ”The fact that all the integrated care is delivered by the same team is a crucial element of FACT. FACT does away with the ‘revolving door’: in good times and in bad, the client has contact with the same team, the same case manager and the same psychiatrist. This is continuity of care” [13]. Even when the clients have been admitted for hospital care, the FACT team is part of the hospital treatment and the decision for release. The team maintains contact with the client during the hospital stay and brings him/her home as frequently as possible, in the daytime or for a few hours. Repeatedly it is pointed out in the manual as an advantage that one single team spans this large field. However, in no place is anything shown suggesting a scientific or experiential basis for such an opinion!

The emergency psychiatric inpatient hospital care operates under completely different assumptions compared to outpatient psychiatric rehabilitation. Possibly, there may be difficulties for the staff to shift between psychiatric care and psychiatric rehabilitation in a credible way, and for this reason it could be that the best result is obtained if there exist professional specialists in both of the areas, specialists who have developed a close and trusting collaboration. If we look at it from the perspective of the client, he or she strives to leave the more hierarchic structure often seen in emergency psychiatric care. That is not the environment one needs when working with improving of one´s abilities and self-confidence. It may seem inhibitory to be forced to have the same professional team around as during ”bad times” when one is set to face a new and better time! If there are built-in difficulties and problems with having one single team for the entire chain of care and rehabilitation, then there is also a basic structural problem with FACT, which in turn might possibly explain the meagre research results.

Such a hypothesis gains support when you read about the so-called Hourglass model in the manual where the ”care workers have to switch roles all the time” [13]. The care and the rehabilitation are described as an hourglass with three separate steps that at times may overlap: (a) Dealing with destabilization, (b) Treatment and (c) Recovery. Judging by the descriptions in the FACT manual, one could suspect that the approach during the destabilization phase might spill over into the other phases. Thus, during Treatment one focuses on the symptoms and one tries to explain the chosen treatment to the client and his/her family and encourage the client to participate in the treatment. It is only during Recovery the client is deemed capable of being placed in the driver's seat, and then the focus is on strength [13].

CONCLUSION

FACT may be seen as a full scale experiment involving a model never tested before: in principle employing one single large team for the entire chain of care and rehabilitation. It has been an audacious and exciting project. Unfortunately, they did not succeed in gaining strong research results and the model now is in need of significant revision. It will most likely lead to a drift toward a traditional ACT or to some other ACT model such as RACT. If such a change will be successful, it is important first to analyze why the FACT model did not fare better.

REFERENCES