preview

Body

Decent Essays

represented within the studies, thereby reducing selection bias with regards to race. Wang et al. (2013) and Holmedahl et al. (2014) took the measurements at a hospital sleep center. Azuma et al. (2014) obtained study subjects from an urban wholesale company in Japan, with the measurements coming from portable monitors worn at the office and at home over the course of one week. Stanchina et al. (2013) used records from hospital outpatients. The remaining studies used measurements taken from university sleep laboratories. Various inclusion/exclusion criteria were applied throughout the studies. Studies such as Holmedahl et al. (2014) and Venkateswaran et al. (2014) intentionally excluded patients with only OSA. Wang et al. (2015) and …show more content…

(2013) (the study with the largest sample size) found a statistically significant difference in the average age of their patient groups (COPD-only, OSA-only and OS). Such differences in demographic data opens the possibility of confounding variables, muddying any analysis. However, other studies had no differences in mean age between the different groups, suggesting that chance or the selection criteria may be playing a large role. Another potential issue is that each study had its own inclusion and exclusion criterion, introducing a risk of bias. For instance, Azuma et al. (2014) only included male subjects in his study, severely limiting the generalizability of its findings. Furthermore, the study also excluded asthmatic patients only by history. Since a post bronchodilator was not used in spirometry testing, asthmatics may have been unknowingly included, thereby possibly clouding the results. Three of the studies included population sizes smaller than 100 people, with Wang et al. (2013) ranking the smallest with 44 test subjects. Small sample sizes increase the risk of chance skewing the results, where the risk decreases with a larger population. Lacedonia et al. (2011) included a fairly small population of 72 people. However, they were broken into five groups: OS, COPD-only, OSA-only, and thin and obese controls. As such, the already small group of 72 people was broken up into subgroups of about 15 each. Once again, with such tiny groups chance is

Get Access