DISCUSSION
This study, which had a relatively large cohort of patients who
underwent surgery for colorectal cancer and were followed up for five
years, showed a relationship between the fulfillment (or lack thereof)
of some accepted diagnostic procedures or treatments employed during
follow-up for this disease with certain adverse outcomes, such as
mortality, recurrence, complications or changes in certain PROM scales.
The yearly performance of CT or colonoscopy in the follow-up of these
patients is related to a lower risk of mortality but, on the other hand,
a higher presence of complications and recurrences as well as further CT
scanning, which may be explained because this type of diagnostic
procedure continues to be used in surviving patients. Regarding PROM
scales, a relationship was also seen with an increase in anxiety and a
decrease in the total EORTC scores, which may reflect a deterioration in
the HRQoL in the evolution of these surviving patients. On the other
hand, the use of chemotherapy was related to higher mortality and
increased presentation of complications or recurrences; radiotherapy, in
this case, was only related to a higher risk of complications, perhaps
due to a smaller sample size in which this treatment was employed. These
treatments are expected to be used in patients with more severe degrees
of the disease, which may explain the higher risk of mortality,
recurrence or complications. In the case of radiotherapy, the four PROM
questionnaires included detected losses in the patients’ health-related
quality of life, while for chemotherapy, losses were found only in the
EORTC-Q30. In any case, losses, as reflected by the beta parameter of
the models, seem to be low losses and, from a clinical point of view, it
seems that the deterioration in their quality of life is minor but the
increase in their chances of surviving is high, but with a greater risk
of complications and recurrences. By using these multivariable models,
we try to show the influence of each of the procedures and treatments
realized at each timepoint of follow-up on the clinical outcomes and
PROMs of the next year of follow-up. Indeed, as different clinical
outcomes and PROMs, as well as the main covariates, changed throughout
the follow-up, an analysis that included all the repeated measurements
and considered the longitudinal design was required to provide
robustness to our results.
Our study tried to correlate just four (but some of the most relevant
and common) quality indicators established for colorectal cancer
patients with specific outcomes 1,10. Although we
attempted in the analysis to control for some of the most relevant
confounders, others could not be included. In any case, it is difficult
to establish a clear cause-and-effect relationship between the study
quality indicators and the outcomes, but we have tried to show that
there are relationships between some of them and that those
relationships are, in these pathologies, complex to interpret since,
performing them are, on the one hand, saving lives, but, on the other
hand, introducing the risk of different adverse outcomes, such as
complications, recurrences or losses in patients’ quality of life11. In summary, the validation of these quality
indices seems difficult to interpret because of complex results;
therefore, the use of these diagnostic procedures or treatments as
quality indices should be interpreted cautiously10,12.