I am a layperson member of ISPS (International Society for Psychological and Social Approaches to Psychosis) who has written previously about how rejecting the “medications for schizophrenia” narrative helped me to recover from severe distress.
In this piece, I wish to critique a recent journal article that illustrates how supposedly “progressive” mental health professionals covertly cling to psychiatry’s disease model.
The article in question is “Antipsychotic treatment: experiences of fully recovered service users,” published in the Journal of Mental Health in March of 2017. The multiple authors of this piece include researchers from Norwegian psychiatric hospitals, Larry Davidson of Yale, and the current Chair of the ISPS Executive Committee, psychiatrist Jan Olav Johannessen. The article purports to be a qualitative investigation of the effects of antipsychotic treatment, one which highlights the perspectives of service users.
While it appears informative and useful, I contend that this article illustrates a far too common reality in psychiatric publishing: a superficial veneer or “iceberg tip” of progressive-appearing research suggesting generalizable new knowledge, covering an “iceberg base” foundation of fallacious illness-model assumptions, pessimistic framings of “mental illness outcome,” and obfuscation of the fact that historical research in psychiatry does not generalize to future settings.
First, there are several genuinely positive aspects to this piece. The authors seriously consider that long term use of antipsychotics may be detrimental, a major step within a profession that has traditionally focused on drug prescribing. They further recommend “presenting antipsychotic treatment as one of many tools supporting the person’s active recovery processes, rather than as the main passive, if not exclusive, way to recovery.” This is an equally progressive view, especially compared to psychiatrists who write about drugging as the primary (and often only) treatment for “schizophrenia.”
Additionally, and in accord with the general public, the article considers psychosocial factors as central to recovery, stating, “Functional recovery beyond the acute phase was, however, primarily attributed to individual efforts and social influences, and less to medication.” And lastly, the authors quote several recovered formerly psychotic people at length — an amazing thing to read, given that most psychiatric articles leave the subjects of treatment voiceless.
This is the tip of the iceberg.
What lies beneath? First, throughout the article, and only briefly interrupted by quotes from service users, the terminology of psychiatry’s disease model permeates what are actually severe problems in living.
Being terrified, nonfunctional, or paranoid is described as “illness” (seven times in the article). How someone feels and functions over time is reduced to “course” (of illness). Giving someone numbing drugs to limit their ability to feel whatever is causing their distress is described as “treatment” (24 times). And general tranquilizers such as Seroquel and Zyprexa are cloaked in the misleading medicalized term “medication” (39 times). The authors could have chosen to use terms like “distress,” “experience,” “neuroleptics,” “administration of oral tranquilizers,” but they do not. Partly as a function of the medicalized language, the distance between the professionals writing the article and the “patients” receiving “treatment” becomes almost palpable.
It seems the authors are unaware that biogenetic language leads to harsher judgments, more prognostic pessimism, increased stigma, less motivation, greater fear, and worse outcomes. Or perhaps they simply do not know another way of writing about their fellow human beings more “humanly.”
I say this seriously: as an outsider, I’ve come to wonder whether medical training and working in hospitals for decades reduces the ability to see one’s suffering fellow human beings as just that. Can a psychiatrist authoring such an article consider that they themselves might become psychotic and have to take tranquilizers if they had been abused or neglected in the way that many of their clients have? Then there is the pressure from journals and one’s fellow professionals to use the ascetic, impersonal, meaningless language of “disease” to describe fear, paranoia, and hallucinations.
While the constant and unevidenced medicalized jargon is problematic, the most damaging quote of the entire piece comes in the introduction:
“Between 9 and 21% of service users with FEP achieve clinical recovery (Jääskeläinen et al., 2013), making this a highly relevant sub-group.”
It is to this faulty assumption that I want to primarily devote this critique. The reader should see that despite positive efforts, the authors mechanistically state — at the very outset of their article — that only 10-20% of the people they, the professionals, label as having “first episode psychosis” (FEP) will recover. This statement is made with very little context regarding FEP — nothing about which service users, receiving which services, for how long, being treated by which professionals, and so on. As is far too often the case with psychiatric journal articles, it is stated as if it is universally true merely because the professionals say it is. As we will see, nothing could be further from the truth.
To make such a pessimistic statement as if it were absolute truth is alarming, resting as it does upon a single meta-analytic study that is severely flawed. One can imagine that other professionals or family members might see such an assertion and decide to stop reading or not seek treatment — after all, why bother trying to understand a problem if the affected person may have only a 10 or 15% chance of getting well?
For those labeled and seeing such a prognostication, the consequences might be even more severe, including an increase of hopelessness and fear and possible suicidal ideation or attempts. Can you imagine how a person feels who is already terrified and isolated, and who after receiving such a diagnosis finds out that professionals think they only have a 10% chance to recover? They may look online — articles such as Johannessen’s are widely available on the internet — and find that authoritative-sounding PhD and MD professionals are extremely doubtful about their prospects, resulting in increased fear and lessening ability to confront the actual sources of their distress.
If these potential harms to service users are real, one has to wonder what goes through the mind of these authors to allow such a statement (an unevidenced one, as I will describe below) to be uncritically published. The statement is so problematic that it deserves a thorough investigation.
The central problem is that the FEP-outcome statement relied on a meta-analysis that contained multiple serious problems.
First, the Jääskeläinen meta-analysis relied on fundamentally flawed assumptions. This study took data from people labeled with “schizophrenia”-spectrum diagnoses (disjunctive labels which are commonly acknowledged to be invalid because people can “have them” while having no symptoms in common), and who were treated primarily at psychiatric hospitals, and frequently given psychiatric drugs. It then suggested that the historical outcomes of this arbitrarily-selected group are what should be expected for future groups of people — under any or no treatment — given such labels by a professional.
The illogicality of these assumptions might not be immediately apparent. But consider the following viewpoint about the generalizability of historical findings in psychiatry by Jan Smedslund in “Why Psychology Cannot Be An Empirical Science”:
“Empirical research in psychology can only be seen as mapping temporarily stable statistical tendencies at the aggregate level. These mappings must be evaluated on pragmatic grounds, i.e. as yielding possibly useful knowledge in limited domains for a limited time… The regularity that is nevertheless observed cannot be taken to reflect permanent laws, but only temporarily stable feedback-loops. These are like whirls in a stream which are stable only as long as the total flow of water does not vary and the stones on the bottom maintain their positions… The findings in psychological journals may superficially appear to be of the same order as those reported in physics or chemistry and, hence, psychology may superficially look like an empirical and accumulative science. The conditional and transient nature of psychological findings is rarely acknowledged, perhaps because the publication and preservation of empirical findings will then appear unjustified since they cannot be taken to be useful in the future. The scarcity of attempted replications also allows psychologists to avoid thinking about this problem.” (emphases mine)
Jääskeläinen therefore took outcomes primarily from psychiatric hospitals employing antipsychotic treatment, occurring under a unique set of past conditions, and assumed that the average of these outcomes represented a universal — rather than conditional — truth about the “course” of the already-invalid label “schizophrenia” (in other words, a mind-twister within a mind-twister, something all too common in psychiatric articles). In using these terms and assumptions, she not only veered far away from describing her fellow human beings in human terms, but also ignored the possibility of much better outcomes achieved in non-psychiatric settings.
Space does not allow me to elaborate here, but please see part two of my article “Rejecting the Medications for Schizophrenia Narrative,” which mentions interventions such as Open Dialogue, 388 program, Soteria, Diabasis, and other forms of intensive psychotherapy that can lead to much higher rates of recovery than 10 or 20 percent.
So what exactly was wrong with the Jääskeläinen study, considered as a support for Johannessen et al’s statement about FEP’s 9-21% recovery rate? In their article, “The Empirical Evidence about Mental Health and Recovery,” Mike Slade and Eleanor Longden thoroughly deconstructed the Jääskeläinen meta-analysis, and I will draw heavily on their critique in analyzing both.
First, it is instructive to consider the seven conclusions that Slade and Longden attributed to professional-leaning, mechanistic approaches to mental distress:
1. Recovery is best judged by experts.
2. Few people with mental health problems recover.
3. If a person no longer meets criteria for a mental illness, they are in remission.
4. Diagnosis is a robust basis for characterizing groups and predicting need.
5. Treatment is needed to improve outcome — and should be widely available.
6. The barriers to receiving effective treatment are availability, financing and client awareness.
7. The impact of mental illness, in particular schizophrenia, is entirely negative.
The authors then contrast these assumptions above (which appear disturbingly close to the assumptions permeating Johannessen et al’s article) with the following seven:
1. Recovery is best judged by the person living with the experience.
2. Many people with mental health problems recover.
3. If a person no longer meets criteria for a mental illness, they are not ill.
4. Diagnosis is not a robust foundation.
5. Treatment is one route among many to recovery.
6. Some people choose not to use mental health services.
7. The impact of mental health problems is mixed.
Slade and Longden then identified three primary problems with the Jääskeläinen article’s pessimistic finding about “schizophrenia” recovery:
1) Sampling strategy, a.k.a. exposure bias — Jääskeläinen primarily took her sample of “schizophrenics” from studies of people entering hospital settings or in regular contact with services. As Slade states, “many people live with psychosis-like experiences outside of mental health services. Their ability to self-manage without attracting the attention of services indicates a lower level of severity and a higher rate of recovery. This means that people with less severe difficulties are systematically less likely to be present in the samples included in the Jääskeläinen review.”
2) Follow-up period, a.k.a. assumption of linearity — Jääskeläinen assumed that the recovery rate of people in the sample was linear, progressing at a fixed rate per year. Slade comments, “No justification is given for this (unstated) assumption, which is undermined by the review finding that duration of followup did not predict recovery estimate.” He continues: “Recovery is not linear. The available empirical evidence indicates that recovery is heavily influenced by context, both social (e.g. social and professional relationships), and psychological (e.g. locus of control, wellbeing)… pooling studies of very different duration into one aggregated analysis is not justified.”
3) Outcome evaluation — The “recovery rates” for people labeled with “schizophrenia” varied massively across the 50 studies comprising Jääskeläinen’s meta-analysis. Slade states: “What recovery rates were found? They ranged across the 50 included studies from 0% to 58%. Even to the casual observer, this might raise some concerns about simply pooling the data to produce a single overall estimate of recovery rate. This problem of combining apples and oranges is known in the systematic review trade as heterogeneity, defined as the extent to which there are genuine differences underlying the results of included studies. Dealing with heterogeneity is a standard challenge in systematic reviews. Are the included studies sufficiently similar to be pooled (or ‘meta-analysed’) to produce an overall estimate of recovery rate?”
Slade and Longden continue by asserting that every key decision Jääskeläinen made leads to more pessimism about outcome — from limiting the study group to those with the most severe problems, to treating studies of markedly different follow-up period as artificial equals, to pooling vastly different studies to produce one global recovery proportion.
Since Slade and Longden wouldn’t say it straight out, I will: the Jääskeläinen study of “schizophrenia outcome” was bullshit. Jääskeläinen’s faulty assumptions about “schizophrenia recovery” parallel the pessimism and equally faulty assumptions that have underlain the fictitious “schizophrenia” diagnosis for decades.
And since Johannessen et al took their numbers about First Episode Psychosis’ recovery rate from such a flawed study, their assertions should also be dismissed. To take such massaged data based on invalid diagnoses as “truth” without acknowledging the severe problems with Jääskeläinen’s meta-analysis, and in equal denial of the far greater recovery rates achieved in multiple alternative approaches such as Open Dialogue, is extremely problematic. It relates to what I will describe below about psychiatry’s generally poor image as an unscientific profession which cannot relate to its “patients” as human beings in their lived contexts.
Ironically, in the “limitations” section of their article, Johannessen et al admit that “Findings are context dependent to the participants and setting in which the study was conducted.” Nevertheless, they seem unaware of the profound contradiction between making this admission and quoting Jääskeläinen’s context-dependent piece, with all its assumptions, as evidence of a supposedly general truth that only 9-21% of people labeled by professionals with First-Episode Psychosis recover.
Considering that one of ISPS’s objectives is:
“Advance education, training and knowledge of mental health professionals in the psychological therapies and psychosocial interventions in the treatment and prevention of psychotic mental disorders for the public benefit regardless of race, religion, gender or socio-economic status.”
One must wonder why one of the leaders of ISPS’s Executive Committee is involved both in uncritically communicating faulty information about this low FEP recovery rate being a general truth, and in using disease-model terminology despite the growing awareness about the harms of using words such as “illness,” “schizophrenia,” “prognosis,” “medication,” “disease,” etc. to describe human suffering.
As a psychiatric survivor my take-home messages, and I direct these squarely at professionals, are the following:
- Stop presuming that severe distress is “an illness.”
- Stop euphemistically referring to general tranquilizers like Zyprexa and Seroquel as “medications.”
- Stop assuming that your definition of recovery is the generally valid one.
- Stop assuming that the rate of supposed “recovery” seen in one quasi-experimental setting under “treatment as usual” is the rate of recovery to be expected for anyone anywhere given the same invalid label. It isn’t.
- And kindly stop labeling us, period.
I am dismayed that in the year 2017, supposedly “progressive” mental health professionals are still writing using the language and assumptions that Johannessen et al employed in their journal article.
As another ISPS member, John Read, wrote in an essay last year:
“A dilemma for all of us who are struggling to broaden our understanding of human distress beyond simplistic, pessimistic, bio-genetic ideology, and to improve our mental health services accordingly, is whether or not to soften our criticisms of psychiatry in the hope of reaching those psychiatrists whose minds are not totally closed. But doing so rests on the assumption that change can come from within the profession. For the last few decades examples of that are few and far between.”
Following from Read, I am somewhat pessimistic that any serious response will be coming from Johannessen et al in response to this critique, even though Jan Olav and I are members of the same group.
But I hope this type of public critique encourages other dissident professionals, along with survivors and family members, to keep speaking up. We should demand immediate changes to the harmful ways professionals label and make pessimistic assumptions around human distress, and we must do so loudly and unapologetically.