|
Home | Issues | eMJA shop | Classifieds | Contact | More... | Topics | Search | Login | Buy full access |
Fiona M Blyth, Ross Lazarus, David Ross, Michael Price, Gary Cheuk and Stephen R Leeder
Electronically published Monday 2 June 1997. Please submit comments by Monday 30 June 1997.
The paper deals with a relatively small sample of patients over a four month period (122 patients based on 88 patients at four months follow-up). The difficulties in identifying cardiac failure from hospital morbidity coding and the problems with assessing quality of life in ill patients are well described. The experience of the investigators in this pilot study are relevant and emphasise the difficulties which will be encountered in larger scale studies.
Detail on the aetiology of the cardiac failure and the criteria for classification of cardiac failure is not provided. This should be noted in the discussion as a point for further large scale studies, but should not delay publication of this pilot study.
[Authors' response: Details of the aetiology have been added in the text. Concerning the criteria for classifying cardiac failure, in a small pilot study of this type it was not feasible to verify the diagnosis of cardiac failure with specialised testing. It is worth noting that most of the study cohort had prior admissions for cardiac failure, which meant that the diagnosis had been established before this study took place. We therefore relied on the clinical judgement of experienced cardiologists and geriatricians working in a teaching hospital. Because of the problems of ICD-9 coding which are discussed in the manuscript, one cannot use this means either for verification of diagnosis. A larger study, with formal verification of the diagnosis and more detailed examination of ICD coding practices is the logical next research step, and we have added a brief comment on this to the Discussion.]
While the sample is small, the investigators have conducted a thorough study and the results are valid, showing that cardiac failure is indeed a heavy burden on patients, hospitals and carers.
It is inappropriate to comment in the Results section on the adverse outcome. If it is necessary to make this point it should be mentioned in the Discussion.
[Authors' response: We have changed the manuscript accordingly.]
The study essentially documents features associated with any admission diagnosis of heart failure. It suffers from having no rigid criteria for confirmation of the diagnosis of heart failure. Essentially, the clinical diagnosis was obtained from records but no clinical criteria are given, nor was there substantiation by investigations such as echocardiography to document structural heart disease. A number of subjects had RNVG, but there is no statement at to whether diastolic dysfunction was detected in any eg. subject with EF of 74%.
[Authors' response: We refer to our first response to referee 1, above. It is worth adding that the research nurses also consulted with medical staff regarding the diagnosis.]
There is also incomplete follow-up of routine subjects, and because of this (and particularly inadequate follow-up in those with extreme poor health) it is almost certain that the burden of heart failure is underestimated long-term.
[Authors' response: We agree that the burden of heart failure is underesstimated in this study. However, in many cases of incomplete follow-up we were able to ascertain that worsening health (or death) played a significant role, and only two subjects were completely lost to follow-up.]
I believe Table 2 and possibly Table 6 could be eliminated. The data in Table 5 could go within the text.
[Authors' response: Tables 2 and 5 have been deleted. What was Table 6 is now Table 2.]
[Authors' response: We agree, and the Table and the text have been changed accordingly.]
2. The last sentence of the abstract needs clarifying or omitting.
[Authors' response: Now omitted.]
3. The information that the authors collected re predictive value of CHF ICD codes could (should) be added to Table 6 [now Table 2]. They should also clarify whether the pulmonary oedemas were coded to 428.1 or other codes such as 518.4 or even 514. These data should be mentioned in Results rather than Discussion.
[Authors' response: We have added some details on this in the Results section - these are necessarily brief due to space constraints.]
4. In the 1st para of Results it needs to be made clear how the 7 missed patients were discovered or how they were missed in the first place.
[Authors' response: This has been clarified in the text.]
5. In keeping with their aim of assessing hospital burden of CHF, some comparison bed-day %ages for other common illnesses might be appropriate somewhere.
[Authors' response: We think this would be interesting to do, but feel unable because of the Journal's space constraints.]
6. On p11, last para there were only 15 deaths in the SF-36 cohort.
[Authors' response: This has been corrected in the text.]
7. Table 5 will also need changing in light of 1. above.
[Authors' response: We have deleted the table and text on the Health Transition Item, which is not usually included in SF-36 analyses, due to space constraints. Further justification for this can be found in reference 14.]
8. Typo in title of Table 2 and last number in Table 3.
[Authors' response: These tables have now been deleted.]
9. I hope the authors intend to examine predictors of the health outcomes of this group, possibly in a further paper.
[Authors' response: This is a good suggestion.]
Please define NYHA grading for CHF.
Tables 2 and 3 could be deleted as the information is explained more clearly in the text.
Indicate from which ethics committee you obtained approval for the study.
Include a box listing your recommendations (i.e., a take-home message for the general medical reader).
[Authors' response: These things have been done.]
Article - Register to be notified of new articles by e-mail - Papers for review list - ©MJA 1997