Bibliographic record
Abstract
Dear Editor, We thank Zhang et al. [1, 2] for their letter and appraisal of our study. The authors describe factors that did appear to be well-accounted for: surgical type, age and body mass index (BMI). Multivariable analysis was performed to adjust for confounding or prognostic effects of such exposure variables from the effect of overall health-related quality of life (HRQOL), which was our primary exposure variable of interest. Age, BMI and extensive operations, such as lobar and oesophageal resections, as well as open approaches placed patients at greater odds of developing complications. Thus, our study adjusted for surgical type, age and BMI. The median age of 66 and a BMI of 28 reflects the patient population we treat at our centre and generally in North America. We did not adjust for anaesthesia type as all procedures were performed under general anaesthesia via endotracheal tube; this is the case for vast majority of cases in Canada and the USA [3]. Despite adjusting for confounding/prognostic variables, it is important not to fall into the Table 2 fallacy; in this regard, we agree future studies should explore subgroups Zhang et al. noted, with sufficient sample size and a-priori design. While post-hoc subgroup analysis was considered, it was ultimately deemed beyond the scope of our study given its exploratory nature and due to the danger of Table 2 fallacy. Additional subgroup analysis on this single-centre sample may also risk alpha inflation and model instability due to increasing inclusion of exposure variables within a limited sample size. Individual HRQOL dimensions were not independently associated with the incidence of postoperative complications. Subgroup analysis according to these HRQOL dimensions was not performed for several reasons, most importantly because a larger sample size is required. Furthermore, our 1st study was meant to assess whether there was any relationship between preoperative overall HRQOL measure and outcomes, which we identified. However, Zhang et al.’s suggestions are astute and can be achieved by 2 means. First, latent class analysis may be performed to identify subgroups of patients who share commonality based on features not directly measurable (i.e. ‘latent’). For example, dimensions such as mood, self-care and mobility could identify clusters of patients similar based on these metrics. Latent clusters of patients may then be compared in relation to their odds of developing complications. This is the analysis we are currently undertaking and will submit for publication in the near future. The second means of confirmation is to repeat this study on a multicentre scale. Increased statistical power would allow traditional subgroup analysis to be performed while decreasing likelihood for type 1 error. Ultimately, meta-analysis of future studies should be performed alongside previous literature describing quality of life and postoperative outcomes [4, 5]. Because many HRQOL tools exist, there remains the risk that the tools will be so disparate as to preclude synthesis unless our community starts to proactively use similar/same HRQOL tools that can potentially be mapped onto each other [6]. Ultimately, by elucidating the utility of HRQOL in identifying high-risk patients, it will allow targeted pretreatment interventions to be administered that reduce risk of postoperative complications. Conflict of interest: none declared.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.007 | 0.064 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.002 | 0.001 |
| Bibliometrics | 0.002 | 0.001 |
| Science and technology studies | 0.002 | 0.003 |
| Scholarly communication | 0.005 | 0.006 |
| Open science | 0.004 | 0.002 |
| Research integrity | 0.028 | 0.040 |
| Insufficient payload (model declined to judge) | 0.007 | 0.008 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".