Maarten van Egmond

56 Chapter 3 This study has some limitations. First, only 155 patients were included in this study who consented to have their functional status measured out of 449 patients treated with esophagectomy between 2012 to 2016 (34.5%). However, the included patients did not substantially differ in baseline characteristics from the overall population (Appendix A). In addition, patients were indicated for surgery only if their preoperative pulmonary function and general condition was good enough for surgery and if metastases were not present. This may explain the high mean values of functional status throughout the pre- and postoperative course. However, the rate of postoperative complications was still high, and there were also a substantial number of patients who scored below normative values of functional status and still returned to baseline levels 3 months after surgery. Second, there was a high number of patients who could not be assessed, mainly during the postoperative course. It could be argued that the patients with postoperative complications or low physical fitness levels were among these patients. However, a detailed analysis revealed that the patients lost to follow-up did not systematically differ in baseline characteristics and functional status at T1. The proportion of patients lost to follow up at T3 with a postoperative complication Clavien-Dindo grade 3a, 4a and 4b was higher compared to the patients measured at T3 and T4, so it could be argued that the postoperative course of functional status may have been different for this group. However, due to the low numbers of patients within each classification grade, it was statistically impossible to analyze the course of functional status of each subgroup. Moreover, severe complications (IIIa and worse) were also present in the patients that completed all measurements. Third, functional status was compared to currently available normative values presented in the peer-reviewed literature. Within physiotherapy, there is a lack of consistency in interpreting these normative values because they have either been based on the means and standard deviations of comparable normative samples or on values from regression equations. The latter should be preferred because it provides continuous norms rather than discrete norms formed by age bands where the consequences of an individual’s raw score can change suddenly as he transfers from one band to another. 37 Another disadvantage of categorizing normative values by demographic characteristics is that it leads to small sample sizes and subsequently wide standard deviations. This could have led to an incorrect interpretation of high versus low functional status. However, these normative values are widely used in physiotherapy practice to indicate functional status and are therefore used in this study as well. In many studies, the importance of determining a patient’s physiological and functional status to successfully recover from surgery has been emphasized. 8,11 Although our study illustrates that functional status returned to baseline values postoperatively and that the course did not