Reliability of the Polygraphic Home Sleep Test for OSA Determined by the Severity and Pattern Changes of Two Consecutive Examinations
Silke Anna Theresa Weber
Round 1
Reviewer 1 Report
Dear authors,
the evaluation of the reliability of poligraphy considering OSA diagnosis and severity is very important. You evaluated 100 patients with possible OSA, who realized 2 sequential home polygraphies, and compared time in bed (TIB), apnea-hypopnea index (AHI), OSA severity, nocturnal apnea pattern, oxygen desaturation index (ODI), and T90 of the two examinations.
I have some questions and suggestions:
Your introduction is very good, describes the problem and the hypothesis.
The objectives are clear.
In the chapter methods:
Did you calculate the size of the study group? this might be inserted, as 100 might not be representative.
Did the patients have diagnostic polysomnography also, to compare to gold standard diagnosis?
I also suggest including which definition of hypopnea you used.
The results are reported in a very confusing form. You compared the severity grades of OSA, however, the composition of these groups changed within the 2 PGs.
all tables are based on comparison of OSA severity degrees, apnea pattern, AHI, thus, they give no real information.
I suggest excluding table 1 and presenting each group's number and demographic characteristics (no OSA, Mild, moderate, and severe OSA).
Table 5: how many had no, mild moderate and severe OSA?
Tabel 6: the exposition is not clear. what do you mean by pattern 1,2,3,4,5,6,sum? pattern changed in 15 patients, 6 changes severity, except severe OSA patients, is this good? discuss the importance of polygraphy for the diagnosis of severe patients,
25% of your group had some changes, that would have an impact on treatment in 40% no OSA, 20% of the mild OSA patients, and 40% of mild OSA patients. It would be interesting to discuss the overestimation of OSA by polygraphy. Which group suffered statistically the most important changes with treatment implications?
You did not show the results of min O2.
In your discussion you cite 2 studies Ahmadi et al who showed 13% of changes considered to be significant, ......your population showed 25% of changes, even worse, please discuss this better. Similar to the study by Gouveriset al who showed changes in 15%.
both studies use different definitions of hypopnea. what was the definition you used? please include in methods.
Your conclusion should answer the question of your objectives, and not repeat results or the justification of the study.
Author Response
Dear Reviewer 1.!
Thank you very much for Your valuable comments and questions!
Please find the answers below!
Your questions:
1.Did you calculate the size of the study group? this might be inserted, as 100 might not be representative.
2.Did the patients have diagnostic polysomnography also, to compare to gold standard diagnosis?
3.I also suggest including which definition of hypopnea you used.
4.The results are reported in a very confusing form. You compared the severity grades of OSA, however, the composition of these groups changed within the 2 PGs.
5.all tables are based on comparison of OSA severity degrees, apnea pattern, AHI, thus, they give no real information.
6.I suggest excluding table 1 and presenting each group's number and demographic characteristics (no OSA, Mild, moderate, and severe OSA).
7.Table 5: how many had no, mild moderate and severe OSA?
8.Tabel 6: the exposition is not clear. what do you mean by pattern 1,2,3,4,5,6,sum? pattern changed in 15 patients, 6 changes severity, except severe OSA patients, is this good? discuss the importance of polygraphy for the diagnosis of severe patients,
9.25% of your group had some changes, that would have an impact on treatment in 40% no OSA, 20% of the mild OSA patients, and 40% of mild OSA patients. It would be interesting to discuss the overestimation of OSA by polygraphy. Which group suffered statistically the most important changes with treatment implications?
10.You did not show the results of min O2.
11.In your discussion you cite 2 studies Ahmadi et al who showed 13% of changes considered to be significant, ......your population showed 25% of changes, even worse, please discuss this better. Similar to the study by Gouveriset al who showed changes in 15%.
Both studies use different definitions of hypopnea. what was the definition you used? please include in methods.
12.Your conclusion should answer the question of your objectives, and not repeat results or the justification of the study
Our answers:
Ad.1.:
The study was based on 100 consecutive patient examinations. The sample number was not calculated. The aim was not to create different subgroups (age, sex, disease profile), so of course, the sample is not representative in this sense. The sample reflects the distribution of the severity of patients admitted at random under normal circumstances
Ad 2.:
We performed a polygraphic test only to interpret how stable the results of two consecutive examinations are in terms of severity and pattern. There have been several studies comparing the reliability and information content of polygraphic and polysomnographic tests (which I cited in the manuscript), but information on repetitive polygraphs is limited.
Ad 3.:
The scoring of the pathological respiratory events was based on the AASM criteria (version 2.6, 2020, 1.B (30% drop, at least 4% of O2 saturation decreasement, pressure flow sensor).
It was added to the text.
Ad 4.:
Sorry for the misunderstanable description.
The first examination represents the patients in a given severity group. The second examination represents the second examination of the same patients. Therefore the composition is the same (same patients with 2 examinations). The two examinations of the same patients were compared with the applied test.
It is clarified in the text.
Ad. 5.:
In our study, we applied the classic parameters used to characterize obstructive sleep apnoea syndrome and to determine the severity of the disease. According to the current guidelines, severity is determined on the basis of the AHI. Both the ODI and the T90 (which indirectly characterize the severity of oxygen desaturation) provide additional information.
The nature and variability of the pattern is judged to provide essential information about the characteristics of sleep apnoea. As far as we know today, the disease has several phenotypes at the clinical level. This is also true for the characteristics of polysomnographic and polygraphic pictures of the disease (pattern), which can be classified into several types and are presumably due to a different pathophysiological background. Examples include the different backgrounds of the supine dependent pattern versus REM dependent pattern. In this respect, the issue of night-to-night stability of the patterns may be important, for which we currently have little data.
Of course, as with any process, the metrics pick out one feature from many aspects of the phenomenon. The most reliable would always be a visual analysis of the raw record and the trend curves of the various parameters under study, but we also need to quantify them from a therapeutic point of view.
Nowadays there are many doubts about the usefulness of AHI as a severity marker of sleep apnoe syndrome, leading to a search for new ways of the description of the disease characteristics (see for example the work of Erna Sif Armadottir).
Ad.6.:
Thank you for Your suggestion but I respectfully disagree with the exclusion of this table. It contains important information about the severity parameters of the different groups. According to Your advice, we complemented Table 1. with the number of elements and minimum oxygen levels of each group.
Ad 7.:
The number of patients in the negative, mild, moderate and severe groups of the first examinations are 8, 29, 27, 36, respectively.
The number of patients in the negative, mild, moderate and severe groups of the second examinations is 10 29, 21, 40, respectively.
The values were inserted into Table 1, which contains the decriptív parameters of the different examined variables.
Ad 8.:
Thank you for Your remark!
The meaning of the numbers of the two axes is the same. I listed the pattern type after the number in the first column. The meaning of the numbers of the row at the top of the figure is the same as in the column on the left side. Sum represents the total number of pattern changes. The number of pattern changes is 15, From these 6 were seen together with severity changes and the more stable patterns were found in the severe category.
We inserted the explanation into the figure legend.
Ad 9.:
Comparing the severity and therapeutic approach changes we demonstrated that the most vulnerable groups are the negative and mild groups. Due to the second examination, therapy change was necessary in these groups only. The second examination did not modify the therapeutic decision in the case of the moderate and severe groups.
On the other hand, false therapeutic consideration would have been made in a few cases on the base of only one examination. Increasement of the severity was fond in 3 cases of negative and 3 cases of mild groups.
The outcome of these data indicates that the negative and mild groups showed the most important changes with treatment implications.
In the above-mentioned two categories, the polygraphic examination can underestimate the real severity of the disease. The therapeutic decision and the possible repetition of the examination must be based on the clinical features of the patient.
It was inserted into the discussion.
Ad 10.:
Instead of the minO2, we described the T90 value (which – together with the oxygen desaturation index – characterizes more detailedly the desaturation background of the disease.
The mean and SD values of the negative, mild, moderate and severe groups of the first examinations are (in %), 87.8, 4.2; 85.6, 4.4; 79.5, 6.4; 68.9, 11.8, respectively.
The mean and SD values of the negative, mild, moderate and severe groups of the second examinations are (in %), 88.5, 2.9; 84.1, 5.7; 77.6, 5.0; 68.3, 11.7, respectively.
We inserted the MinO2 values into Table 1.
Ad 11.:
In Ahmadi's work, the difference between the two overnight tests was significant if the change in AHI between the two was greater than 5/h. It doesn’t represent real severity changes in all cases.
In contrast, in our study, we used clinically significant change (associated with a change in therapy and a change in formal severity grade).
In Ahmadi's work, a change was described in 13% of patients in the second study, whereas in our study, only 6 out of 100 cases had a change in severity grade (6%) that warranted a change in therapy.
We consider that clinically significant change (requiring a change in therapy) is a more useful approach for daily practice.
Thank you for Your remark.
It was inserted it into the methods!
Ad 12.:
Than you for Your remarks!
The conclusion is rewritten according to Your suggestions!
Thank you very much again:
Béla Faludi
Author Response File: 
Author Response.pdf
Reviewer 2 Report
The concept of current study is interesting and worthy. However, there are several concernss, especially in study design.
Major
-100 subjects were enrolled in this study. What is the inclusion/exclusion criteria?
Is this a prospective study? If it is, how the number of enrolled subjects were calculated?
Authors should described the duration of the study and participants enrollment.
Is there no one who did not finish the whole study, and dropped out during the study?
Methods section should be complemented.
-Authors should add information about clinical characteristics of enrolled subjects.
How about the alcohol habit? Did author have concurrent medical diseases such as cardiovascular disease? How about the medication history of enrolled subjects?
Please suggest the characteristics of enrolled subjects including clinical and polysomnographic things as a dependent table.
-Did the age, gender, alcohol consumption, or medical history affect the correlation between two sleep tests?
The effect of clinical characteristics on correlation between two sleep tests need to be further analyzed in Results and Discussion section.
Author Response
Dear Reviewer 2!
Thank you very much for Your valuable comments and suggestions!
Please find our answers below:
The questions were:
- 100 subjects were enrolled in this study. What is the inclusion/exclusion criteria?
- Is this a prospective study? If it is, how the number of enrolled subjects were calculated?
- Authors should described the duration of the study and participants enrollment.
- Is there no one who did not finish the whole study, and dropped out during the study?
- Authors should add information about clinical characteristics of enrolled subjects.
- How about the alcohol habit? Did author have concurrent medical diseases such as cardiovascular disease? How about the medication history of enrolled subjects?
Please suggest the characteristics of enrolled subjects including clinical and polysomnographic things as a dependent table.
- Did the age, gender, alcohol consumption, or medical history affect the correlation between two sleep tests? The effect of clinical characteristics on correlation between two sleep tests need to be further analyzed in Results and Discussion section.
Our answers:
Ad.1.:
Altogether 100 patients were enrolled in the study. The patient group was unselected. During a routine polygraphic test for sleep apnoe syndrome the first 100 consecutive patients were selected for further analysis, independently from the symptomatology and other cardio- and cerebrovascular diseases. The inclusion criteria were the two technically good polygraphic examinations on two consecutive days. Patients with less than six hours of recording and a lack of two examinations were excluded from the study.
Participants were instructed to do everything in the same way (sleep timing, eating habits, medication) as on the nights without tests.
The patients were asked to avoid alcohol consumption during the examinations. All the patients filled out a questionary about their former medical history and medication including cardiovascular diseases.
Since two consecutive nights of testing were performed, the possible effect of the disease profile changes on the result of the examination was excluded.
Details added to the text!
Ad 2._
Our study is not a prospective study. 100 unselected, consecutive patients were enrolled to the study, without follow-up period. Only two examinations were done in two consecutive days.
Ad 3.:
The enrollment period was between 1st of July, 2020 and 15 of September, 2021.
Inserted to the text.
Ad 4.:
Only that patients were enrolled to the study who underwent the two polygraphic examinations on two consecutive days. Therefore the patient with one polygraphic examination (or examintion with technical problems) did not fulfill the inclusion criteria of the study. The number of participants excluded from the study is 23.
Added to the text.
Ad. 5.:
Thank you very much for YOur advice!
The disease profile of the participants is summarized in Table 7.
Ad 6.:
Study participants completed a questionnaire about co-morbidities, medication and alcohol consumption habits.
Only occasional alcohol consumption was indicated in the questionnaires and patients were instructed not to consume alcohol during the study. Thus, alcohol consumption habits did not influence the study results
The clinical and polygraphic characteristics of the participants were summarized in Table 7.
Table 7 shows a trend that, as the severity of the disease increased, the proportion of different types of cerebro- and cardiovascular diseases increased in the study population.
It is inserted into the text.
Ad 7.:
The goal of our study was only to compare the result of the two polygraphic examinations.
Participants were instructed to do everything in the same way (sleep timing, eating habits, medication) as on the nights without tests. This best represents the condition to which the therapy should be adapted.
Since two consecutive nights of testing were performed, changes in the disease profile and its consequential effect ont he examination results were not expected.
Study participants completed a questionnaire about alcohol consumption habits. Only occasional alcohol consumption was indicated in the questionnaires and patients were instructed not to consume alcohol during the study. Thus, alcohol consumption habits did not influence the study results
In the present study, the correlation of the two tests with the age, sex and comorbidities was not explored as the number of possible groups would require a much larger number of items. Further studies along these lines are planned in the future and data collection is ongoing.
A note on this has been inserted in the text.
Thank you very much again:
Béla Faludi
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report
The authors clarified the requested points in the text, improving the article. thank you
Author Response
Dear Reviewer 1!
Thank you Your suggestions.
The spellcheck was made.
Best regards: Béla Faludi
Reviewer 2 Report
Authors tried to improve their manuscript following reviewers' suggestion.
Although it is not completely designed, the study itself seems to be qualified.
Author Response
Dear Reviewer 2!
Thank you Your suggestions!
The result was modified, substituted.
Best regards: Béla Faludi
