Preoperative rehabilitation for patients undergoing colorectal cancer surgery: a retrospective cohort study
- PDF / 177,481 Bytes
- 2 Pages / 595.276 x 790.866 pts Page_size
- 37 Downloads / 185 Views
LETTER TO THE EDITOR
Preoperative rehabilitation for patients undergoing colorectal cancer surgery: a retrospective cohort study Takaaki Ikeda 1,2
&
Toru Tsuboya 2
Received: 5 September 2019 / Accepted: 14 October 2019 # Springer-Verlag GmbH Germany, part of Springer Nature 2019
We read the manuscript with great interest, which is written by Watanabe and the colleagues, entitled “Preoperative rehabilitation for patients undergoing colorectal cancer surgery: a retrospective cohort study” [1]. The authors analyzed 1934 patients from more than 100 acute-care hospitals across Japan and examined the association of preoperative rehabilitation on activities of daily living (ADL) and complications among patients with stage 1 and 2 colorectal cancer who underwent surgery. The authors concluded that short-term preoperative rehabilitation was associated with better ADL and fewer complications [1]. This study is of importance to discuss whether short-term preoperative rehabilitation is clinically important for patients with colorectal cancer who are going to receive surgical treatments. However, we have several concerns about their methodology and interpretation of the study findings as bellows:
Reliability of the data and generalizability of the findings We have concerns about the reliability of the data and the analysis. First, the length of hospital stay (LOH) among patients in Watanabe’s paper seems to be much longer than previous studies of colon cancers conducted in Japan [2] and other countries [3, 4]; LOH was 24–26 days in Watanabe’s study (study duration was in 2014–2017), while LOH was 7– 22 days in other studies (study duration was in 2003–2013) [2–4]. Thus, we doubt their findings can be applicable to general patients. Obviously, we understand LOH is influenced by * Toru Tsuboya [email protected] 1
Department of Health Policy Science, Graduate School of Medical Science, Yamagata University, Yamagata, Japan
2
Department of International and Community Oral Health, Tohoku University Graduate School of Dentistry, 4-1, Seiryo-machi, Aoba-ku, Sendai, Miyagi 980-8575, Japan
various clinical factors and could vary depending on situations, but the huge gap in LOH between Watanabe’s paper and the other related papers should be discussed. Second, the sample size increased from 1934 to 2370 when they used the IPW method. In our understanding, the sample size would NOT be changed if the IPW method was applied. We suspect they used “frequency weighting” method rather than the “analytic weighting” one, and if so, we have concerns about falsepositive results, due to increased sample size [5].
Study design In Watanabe’s paper, they used two outcomes: (1) the proportion of patients who showed the Barthel Index (BI) decline during hospitalization and (2) the number of complications during hospitalization, assessed using ICD-10 codes. We consider those are insufficient to reflect clinical importance. Firstly, we believe we need to use absolute BI score difference itself, not “the proportion of patients who showed Barthel I
Data Loading...
