Comment on "The effect of preoperative chest physiotherapy on oxygenation and lung function in cardiac surgery patients: a randomized controlled study".

We read the article titled “The effect of preoperative chest physiotherapy on oxygenation and lung function in cardiac surgery patients: a randomized controlled study” published in your journal.[1] I would like to commend the authors for evaluating the effect of preoperative chest physiotherapy in 100 patients undergoing cardiac surgery.

The article is plausible and complete in every aspect, however, there are a few other statistical tests and crucial reporting items missing. First, about the statistical test, it would be better to compare the mean difference between the two groups or use ANCOVA with baseline measurements as covariant.[2,3] By doing so, authors will be able to provide robust evidence to use pre-operative physical therapy modalities in patients undergoing cardiac surgery.[4] Second, authors need to report the effect size for readers to evaluate the magnitude of the experimental effect.

Reply

We would like to thank the readers for their valuable and constructive comments regarding our published study entitled “The effect of pre-operative chest physiotherapy on oxygenation and lung function in cardiac surgery patients: a randomized controlled study”.

We take your feedback open-heartedly but we had not previously performed the ANCOVA test because we found that all the patients in our article were matching in the baseline and their results were comparable, so we compared them without comparing each patient individually. However, we appreciate learning from your wisdom and experience. So we ran the ANCOVA test as required and report the results as follows :

Analysis of the intervention effect on FVC (%predicted) using ANCOVA test revealed a statistically significant difference between the intervention (group I) and the control group (group II) in the postoperative (day 7) FVC (%predicted) after adjusting the mean difference. A pairwise comparison to estimate the effect size revealed that the intervention group (group I) had a greater postoperative FVC (%predicted) than the control group (group II) of 21.96 (95% CI, 18.96-224.97) ().

Analysis of the intervention effect on FVC using ANCOVA test revealed a statistically significant difference between the intervention and the control group in the postoperative (day 7) FVC after adjusting the mean difference. A pairwise comparison to estimate the effect size revealed that the intervention group (group I) had a greater postoperative FEV1 (% predicted) than the control group (group II) 22.12 by (95% CI, 19.2-25.0) ().

We found that the conclusion was unchanged. Based on the primary outcome of the study and calculation of Cohen's d, the effect size was larger than 1 and hence clinically important.

Table 1.

Analysis of the intervention effect on FVC using ANCOVA test.

Source	Type III sum of squares	df	Mean square	F	Sig.
Corrected model	18481.981a	2	9240.991	166.415	<.001
Intercept	6605.974	1	6605.974	118.963	<.001
Basal FVC (% predicted)	4364.842	1	4364.842	78.604	<.001
Group	11695.231	1	11695.231	210.612	<.001
Error	5386.379	97	55.530
Total	592686.000	100
Corrected total	23868.360	99

Table 2.

Analysis of the intervention effect on FEV1 using ANCOVA test.

Source	Type III sum of squares	df	Mean square	F	Sig.
Corrected model	20110.029a	2	10055.014	197.808	<.001
Intercept	5832.981	1	5832.981	114.750	<.001
Basal FEV1	4741.117	1	4741.117	93.270	<.001
Group	11631.188	1	11631.188	228.815	.<001
Error	4930.721	97	50.832
Total	634221.000	100
Corrected total	25040.750	99