Title: The psychosocial difficulties in brain disorders that explain short term changes in health outcomes

Transcription

1 Author's response to reviews Title: The psychosocial difficulties in brain disorders that explain short term changes in health outcomes Authors: Alarcos Cieza Cristina Bostan Jose Luis Ayuso-Mateos Cornelia Oberhauser Jerome Bickenbach Alberto Raggi Matilde Leonardi Eduard Vieta Somnath Chatterji Version: 6 Date: 24 September 2012 Author's response to reviews: see over

2 Prof Pim Cuijpers Editorial office BMC Psychiatry Revision of manuscript: Are psychosocial difficulties common across brain disorders? Dear Editor, Thank you very much for the reviewers comments on the above mentioned manuscript. We found them very helpful. The manuscript has been revised, taking your comments and the comments of the reviewers into account. Please find bellow a detailed description of the corresponding amendments and changes. The words and sentences that have been deleted are crossed and the words and sentences added underlined. You also find attached revised version of the manuscript. Yours sincerely Professor Alarcos Cieza Faculty of Social and Human Sciences, University of Southampton, Highfield Campus, Southampton SO17 1BJ United Kingdom

3 COMMENTS OF THE EDITOR 1. Please give the full names and affiliations of the ethics committees that approved the study. Please add this information in full to a revised manuscript. The full names and affiliations of the ethics committees have now been included in the revised manuscript. The corresponding paragraph has been changed as follow: The study protocol and the informed consent forms were approved by the responsible Ethics Committees in at each center., namely Comité Ético de Investigación Clínica, Hospital Universitario de la Princesa, Madrid, Spain, Comitato Etico Dell Istituto Nazionale Neurologico Carlo Besta, Milan, Italy, Eticka Komise Vseobecne fakultni Nemocnice v Praze, Prague, Czech Republic, and Institut Republike Slovenije za Rehabilitacijo, Ljubljana, Slovenia. We also attached the corresponding approvals for review by the editior 2. - Copyediting: After reading through your manuscript, we feel that the quality of written English needs to be improved before the manuscript can be considered further. We advise you to seek the assistance of a fluent English speaking colleague, or to have a professional editing service correct your language. Please ensure that particular attention is paid to the abstract. The manuscript has been read by an English native speaker with experience in the field of brain disorders and knowledgeable of the ICF. The abstract has also been improved. 2

4 COMMENTS OF REVIEWER: Annet Kleiboer 1) The title should correspond better with the content of the paper We want to thank the reviewer for this and the following comments that have helped us to realise that there is no consistency between the title, background research question, analyses, discussion and conclusion. As suggested by the reviewer, we have accordingly changed title, parts of the background, the discussion and the conclusion. The title has been changed as follows: The psychosocial difficulties in brain disorders that explain short term changes in health outcomes The other changes are described in the following answers. 2) The background, discussion and conclusion of the paper do not correspond with the research question and the way the data was analysed. The background and conclusion suggest that the authors are interested in identifying shared psychosocial difficulties and their relationship with health outcomes in people with a wide variety of brain disorders. The research question and analyses focus on the relationship between a set of general psychosocial difficulties and their relationship with health outcomes across a wide variety of brain disorders. The research question and analyses do not provide information on shared psychosocial difficulties across disorders. The authors should match background, research question, analyses, discussion, and conclusion. We agree with the reviewer and have realised that talking about shared psychosocial difficulties in brain disorders introduced confusion into the paper and that identifying shared psychosocial difficulties is not the objective of the paper. To make background, research question, analyses, discussion and conclusion consistent, the following changes have been made: 1) Sentences of the background discussion and conclusion that make reference to the identification of common psychosocial difficulties across brain disorder have been deleted. 2) The rational of the study has been clarified by adding sentences that explicitly mention the need to identify psychosocial difficulties that explain health outcomes and why the ICF can be a useful tool to achieve the aim of the study. Bellow you find the changes that have been made to the introduction, discussion and conclusion. INTRODUCTION Paragraphs 1 to 3 were not changed. The remaining paragraphs were changed as follow: There are some few European studies that include descriptions and assessments of psychosocial difficulties associated with brain disorders. These studies, however, tend to focus on single brain disorders, or simple combinations of one or two of them. As a result, the available information on psychosocial difficulties takes the form of narrow information silos that are neither comprehensive nor comparable across disorders, without extensive attempts at post hoc harmonization (Error! Bookmark not defined.). It is also the case that up to now most studies that address psychosocial difficulties tend to use a cross-sectional rather than a longitudinal design. This may be because psychosocial difficulties 3

5 associated with brain disorders tend to follow a chronic trajectory, in which changes over short periods of time are small. Even so, small changes in psychosocial difficulties and global outcomes may well reveal important intervention targets that otherwise would be ignored, to the detriment of the patient. Some of these studies have used tthe WHO s International Classification of Functioning, Disability and Health (ICF) (9) has been shown to be an appropriate tool for describing as framework for describing psychosocial difficulties of brain disorders, which in the ICF are called disabilities (10, 11, 12, 13, 14). Since the ICF is an etiologically neutral classification, it has the great advantage that can be used for describing the psychosocial difficulties of any brain disorder. Based on the need to carry out studies that go beyond the common practice of focusing on a single brain disorder, or simple combinations of one or two of them (Error! Bookmark not defined.,15), There is reason to believe, therefore, that our clinical and epidemiological our understanding of psychosocial difficulties associated with brain disorders can take advantage of could profit from using the ICF as a basis for systematically describing to structure a systematic description of these difficulties. In addition, the ICF also includes environmental factors understood as determinants that can have a potential positive or negative influence on health outcomes. The use of the ICF for data collection is thus further justified as it allows for the identification of those environmental determinants. The ICF is etiologically neutral in the sense that the disabilities a person experiences are directly attributed to any underlying health condition or disease, but rather are understood as outcomes of interactions between intrinsic health features of the person and environmental and personal factors: two people with very different diseases may have the same kind and level of disability, and two people with the same disease may experience different kinds and levels of disability. There is also a need for studies that address psychosocial difficulties based on a longitudinal design. Our clinical experience teaches us that the psychosocial difficulties of persons with brain disorders fluctuate with the stage of the disease, not infrequently over short periods of time. It is common, for example, to observe changes in the levels of energy and drive, fluctuations in emotionality and in the extent in which the work performance is affected by the disease. An open question so far is the extent to which those short-term fluctuations at any stage of the disease process affect general health outcomes. This kind of information would be of There would therefore be considerable value for interventions and health systems planning if it turned out that diverse brain disorders shared a range of psychosocial difficulties. This is an empirical question that requires investigation. The aim of this study is, therefore, to identify a set of psychosocial difficulties or disabilities using the ICF that are associated with and determine short term changes in health outcomes across heterogeneous brain disorders. The following reference has been added to the introduction: Wittchen HU, Jacobi F, Rehm J, Gustavsson A, Svensson M, Jönsson B, Olesen J, Allgulander C, Alonso J, Faravelli C, Fratiglioni L, Jennum P, Lieb R, Maercker A, van Os J, Preisig M, Salvador-Carulla L, Simon R, Steinhausen 4

6 HC. The size and burden of mental disorders and other disorders of the brain in Europe Eur Neuropsychopharmacol Sep;21(9): DISCUSSION The first sentence was changed as follow: This study identified a set of psychosocial difficulties common across heterogeneous brain disorders that are associated with and determine short term changes in health outcomes in a group of heterogeneous brain disorders. The first sentence of the third but last paragraph was changed as follows: Finally, this study shows that it is useful to use the ICF as a source of independent variables for longitudinal studies and as a basis for defining psychosocial difficulties. Using ICF made it possible for us to analyze together the data derived from different brain disorders, thus developing what might be called horizontal epidemiological results. CONCLUSION The first sentence was changed as follow: Our findings are of primary interest to researchers and clinicians as they summarize a comprehensive list of relevant psychosocial difficulties that, for the first time, have been shown to be common to a broad range of brain disorders and explain the variability in health outcomes. 3) Rationale: In the background, the authors criticise previous studies for focussing on specific brain disorders or simple combinations of one or two of them. It is not clearly described why the authors have combined a wide variety of brain disorders (both psychiatric and neurologic). It should be made clear in the introduction (1) what the rationale is for combining such a wide variety of common brain disorders, (2) what potential overlap regarding psychosocial difficulties the authors expect between the different brain disorders and why, (3) what differences they expect and (4) what the differences may be with other (non-brain) disorders, to provide a more solid background for the study. As mentioned, we agree with the reviewer that it was very confusing to talk about shared psychosocial difficulties and that there was no need to do so. As can be seen from the changes presented in the answer to the previous comment we have deleted the criticism of previous studies. We have not made additional changes since these comments were addressed by deleting all sentences in the paper that talked about shared psychosocial difficulties, as explained in the previous answer. 4) Background: The authors were interested in predicting a change over time in health outcomes from a set of psychosocial difficulties. It is not clear why they would expect a difference in the 12 weeks following baseline. The courses of the included diseases are very different and the participants were not recruited at a specific moment in their disease trajectory. For example whereas psychosocial difficulties may predict health outcomes shortly after stroke, there is no obvious reason why they would predict health outcomes in migraine. Thank you for the comment that helped us realise that we had to be more explicit why we have followed up health outcomes and psychosocial difficulties over 12 weeks. 5

7 To address this issue, we have changed the last but one paragraph of the introduction. This now reads as follows: There is also a need for studies that address psychosocial difficulties based on a longitudinal design. Our clinical experience teaches us that the psychosocial difficulties of persons with brain disorders fluctuate with the stage of the disease, not infrequently over short periods of time. It is common, for example, to observe changes in the levels of energy and drive, fluctuations in emotionality and in the extent in which the work performance is affected by the disease. An open question so far is the extent to which those short-term fluctuations at any stage of the disease process affect general health outcomes. This kind of information would be of There would therefore be considerable value for interventions and health systems planning if it turned out that diverse brain disorders shared a range of psychosocial difficulties. 5) Methods (design) section: The follow-up period for the study was set at 6 weeks and 12 weeks after baseline but the time points were irregularly spaced and the actual times in days from baseline varied widely ( days for T1 and days for T2). You can control for this in the analyses but it remains a difficult problem to deal with. In addition I expect that the delays in follow-up are not random (ie I expect that time lags may be related to specific brain disorders or patient characteristics). The authors should explain more clearly why this is not a threat to the validity of the results? We agree with the reviewer that the time points were irregularly spaced. Our methodology accounts for these differences in time points, since time in days is included as a metric variable in the multilevel models for change. Nevertheless to address this comment we decided to exclude from the analyses the data of those patients who were outliers in the time points of assessment. We excluded the values of T1 that were <21 days and > 84 days that is, half of the 42 days to twice the 42 days after T0, when T1 should have taken place. We also excluded those values of T2 <42 and >168 half of the 84 days to twice the 84 days after T0, when T2 should have taken place. The corresponding sentence of the Methods section has been changed as followed. Data on T1 was collected between and days after baseline with a mean of and a median of 42 days, data on T2 between and days after baseline with a mean of and a median of 88 days. The time points of assessment are much more regularly spaced after deleting the values of outliers. In addition, as mentioned before, since time in days is included in the model as a metric variable the validity of the results is not threatened by this variability. Since some outliers were excluded, the whole analyses have been performed again. The values in the tables have slightly changed. We have correspondingly changed the tables and the results section. The discussion was also slightly changed. 6) Recruitment procedures should be described better (where recruited, how, when). Thank you very much for this comment. We have added the following sentence has been added to the first paragraph to the section Data collection procedures : 6

8 Consecutive patients in the six study centers mentioned above were recruited from September 2005 to May 2008 by the trained health professionals, who were psychologists, neurologists, psychiatrists and rehabilitation doctors. 7) Results: If the aim is to identify a set of shared psychosocial difficulties across brain disorders it would be necessary to compare health outcomes and psychosocial difficulties between diseases. As mentioned before, we agree with the reviewer that it was confusing to talk about shared psychosocial difficulties and that there was no need to do so. Based on the changes made to the manuscript that reflect that the aim of the study was NOT to identify a set of shared psychosocial difficulties, we do not show the results of the analyses suggested by the reviewer. 8) The discussion and interpretation of the results should be adjusted according to the comments mentioned above (especially comment 2-4). Please see the answers to the comments 2 to 4. We have performed the corresponding changes to the discussion. 7

9 COMMENTS OF REVIEWER: Ulrich Hegerl 1) Methods Study design: There is a high overlap of T1 and T2 (T1: days; T2: days). This fact should be mentioned as a methodological limitation. Moreover, it will be interesting for the reader to know the reasons for this extremely high range of days. As mentioned in the comments to the first reviewer, we have excluded from the analyses the data of those patients that were outliers in the time points of assessment. The corresponding sentence of the Methods section has been changed as followed. Data on T1 was collected between and days after baseline with a mean of and a median of 42 days, data on T2 between and days after baseline with a mean of and a median of 88 days. In addition, to address the concern of this reviewer, we have added to the last sentence before the conclusion the following sentence: A second limitation of the study that should be mentioned is the irregularity with which the time points of assessment are spaced. This irregularity was due to practical considerations, such as the time in which the health professionals could access the patients clinical records and performed the interviews. Nevertheless, the irregular time points of assessment across patients does not put into question the validity of the results, since time in days has been included as a metric variable in the multilevel models for change. As mentioned in one of the answer to the first reviewer, since some outliers were excluded, the whole analyses had to be performed again. The values in the tables have slightly changed. We have correspondingly changed the tables and the results section. There was no need to change the discussion. 2) Methods Measures: Page 9, line 1: The authors mention that the diagnosis was based on International Classification of Diseases (ICD)-10 criteria. Were the diagnoses verified clinical diagnoses or based on a structured clinical interview? The diagnoses were verified clinical diagnoses. We have added this information in the corresponding part of the Study design section. This is the change made: our study was a longitudinal observational investigation on a convenience sample of persons with the verified diagnosis of one of the following brain disorders: 3) Methods Measures: Beck Depression Inventory II: Why did the authors decide to use a self-report instrument for the assessment of depression? It would have been more convincing to use an assessment tool by physicians like the Hamilton Depression Rating Scale. Thank you very much for this comment which addresses a very important issue that was intensively discussed by the research team planning the project. The reasons why we decided using the Beck Depression Inventory were twofold. Frist, for practicability since the Hamilton Depression Rating Scale was not routinely documented from all the patients and secondly because, though there are some disadvantages using patient-reported scales, such as overestimation of symptoms severity, there is also evidence that confirms that physician-report scales, such as Hamilton Depression Scale, might not properly describe patient's clinical situation, especially in depressive patients who may be in remission (Zimmerman 2012). 8

10 We have not added any sentence in the paper addressing this issue because doing so would give to the description of the instrument selected for depression more importance than to the other instruments selected for other conditions. This extra emphasis on depression would be unjustified. Zimmerman M, Martinez J, Attiullah N, Friedman M, Toba C, Boerescu DA. WHY DO SOME DEPRESSED OUTPATIENTS WHO ARE NOT IN REMISSION ACCORDING TO THE HAMILTON DEPRESSION RATING SCALE NONETHELESS CONSIDER THEMSELVES TO BE IN REMISSION? Depress Anxiety Aug 7. doi: /da [Epub ahead of print] 4) Results: The fact that 103 ICF categories had to be excluded from the analyses should be mentioned as an important bias in the discussion. Thank you again for this comment. We agree that it is important to address this issue in the limitations of the study. Thus, the following sentences have been added to the paragraph before the conclusion: Finally, 103 ICF categories had >50% of missing values and were excluded from the analyses. As mentioned before, we followed the recommendations of WHO and only those ICF categories that are relevant to the patient are documented. Thus, we assume that the ICF categories excluded from the analyses refer to issues that are not relevant for the large majority of the patients and therefore the validity of the results are not compromised. However, we cannot deny that the decisions of the health professionals regarding the relevance for the patients of areas addressed by ICF categories is fallible, which might have influenced the validity of the data at the end. For future studies, we recommend that the same ICF categories always be evaluated for all the patients included in the study. 5) Results Page 14, lines 18-19: Age was a significant predictor just for WHODAS II score, with higher age being related to worse health. According to Table 4, there is a negative association between age and the WHODAS II score suggesting that higher age is associated with better health (estimate: -2.27; p <0.05)! Thank you for this comment. The sign of the estimate mentioned by the reviewer was a mistake. We had to change the Results section and the tables because the complete analyses had to be performed after deleting some extreme values regarding the duration between time points. The association between age and the WHODAS II score is significant and positive as one would expect. All the results in the table have been checked several times to make sure that there are no mistakes in the data presented. 6) Results Pages 15-16: The authors report in detail which ICF categories were significantly associated with health outcomes in the sample. However, they do not specify the direction of the associations. For example, it would be important to note that sleep problems were significantly associated with worse mental functioning. We have addressed the comment of this reviewer by adding always the direction of the association in the Results section. 7. Discussion: The PARADISE project should be described with more detail. 9

11 The following sentences describing the PARADISE project have been added to the corresponding part of the discussion: PARADISE is currently testing an innovative approach to collecting clinical data on the psychosocial difficulties that people experience when they suffer from brain disorders, the determinants of the occurrence of those difficulties and the determinants of their change over time. The results of this project will be available mid Minor changes: # References 17 and 24: Please substitute Ustun by Üstün. This minor change has been made. 10

12 COMMENTS OF REVIEWER: Lukasz Cichocki 1) The form of results' presentation is very unclear. The most important table table number 4 - is too vast and complicated. It should be divided in at least 3 parts. Taking into account the size of the article this decision leads to conclusion that maybe authors should choose some of the resluts and not present all of them in one paper. Thank you very much for the comment. We have added the following footnote to Table 4, to clarify its interpretation: Column Estimate contains the mean of the regression coefficients and the column SE the standard error of those estimates. The estimates have to be interpreted with respect to the reference values, which were male for gender, stroke for the brain disorder, and high for disease severity. For the ICF categories the reference value is not having a problem. While positive estimates in the GH and MCS indicate better health, they indicate lower health in the WHODAS. The SE has to be interpreted in such a way that the smaller the SE, the more precise is the estimation of the regression coefficients. We hope that given changes already described, this reviewer would agree that there is no longer any need to present some of the results in another paper, what we think would weaken the scientific value of this paper. 2) Presentation of results is also unclear in the text. Too many abbreviations, symbols e.g. regarding questions in questionnaire. Thank you very much for this comment. To improve the readability of the paper we have deleted the codes of the ICF categories. Only their titles are now mentioned. To make sure that it is clear that we refer to ICF categories, these titles have been marked in italics. We have added the following sentence in the paper to make this clear to the reader: To facilitate the reading of the paper, we will not use the codes of the ICF categories but only their title from now on. 3) The language of this article is very hermetic and not accesible to many potential readers. It is rather like the language of a European Commision document. Authors should pay more attention to whom this aricle is addressed and underline a more practical purpose and sense of what they have described. Thank you for this comment. As mentioned to the editor, the paper has been read and corrected by an experienced English writer who knows the field. We very much hope that the readability of the paper would now satisfy this reviewer. 11