Clinical Judgment in the Provision of Optimal Care vs Selection Bias in a Survival Analysis

Maurie Markman, MD

Clinical judgment is an ill-defined but well-appreciated factor that is highly likely to affect the provision of optimal care for patients with any serious medical condition, including those with cancer.

In the realm of surgical oncology, surgeons rely to a major extent on their clinical judgment when making decisions, such as whether a 78-year-old man with a prior myocardial infarction can withstand the rigors of a curative resection of colon cancer with a single localized metastatic lesion in the liver, or whether a woman with newly diagnosed intra-abdominal carcinomatosis from a presumed ovarian primary should undergo primary cytoreductive surgery or a neoadjuvant approach.

Similarly, the potential benefits vs risks of retreatment of a previously irradiated area to provide palliative pain relief depend on the technical knowledge edge and experience of the consulting radiation oncologist. Decisions on chemotherapy, drugs, dosages, and schedules, which mandate a thorough understanding of the available evidence-based data as well as patient goals, and the ability to satisfactorily communicate and appropriately modify suggested actions based on situational changes are all relevant components of effective clinical judgment.

Finally, it is doubtful there will be much, if any, disagreement with the statement that sound clinical judgment is an important feature characterizing the most effective cancer physicians.

Yet, a problem arises when we pivot from that decision-making process for individuals, which almost certainly occurs multiple times during a given patient’s cancer journey, and attempt to specifically ascribe a favorable or unfavorable objectively measured outcome (eg, survival, development of locally recurrent disease) to a particular therapeutic strategy that has been selected at least partially based on that judgment.

Clearly, there are settings where an outcome may be directly influenced by the skills and prior experience of an individual clinician and the institution, as well as the resources available to optimize that care.

An incident from early in this commentator’s career related to the interpretation of results from a preoperative radiation therapy study in rectal cancer emphasizes this point. This study revealed the benefits associated with the delivery of radiation prior to surgery compared with surgery alone in reducing the risk of local disease recurrence. I vividly remember asking a surgical colleague at the time of the study’s publication whether he felt these results would influence his choice of management. He replied that it was his opinion that the surgeons who participated in the randomized trial required the assistance of local radiation because of a less-than-optimal technique that led to an unacceptably high recurrence rate. For the record, I was also provided with institutional data to support this claim.

Although much has changed from this time in the realm of colorectal surgery, radiation, and imaging technologies, the point to emphasize is the relevance of clinical skill, experience, and individual judgment in cancer outcomes.

However, as important as clinical judgment is in influencing optimal cancer care, it would be an error to conclude that it is the only or even necessarily the most significant factor in ultimately determining clinically relevant outcomes.

To emphasize this point, consider the simple example of one of the many components of care that might influence treatment outcomes: the distance patients have traveled to receive their care. Although the impact of distance may be a negative prognostic feature in the absence of satisfactory local options (eg, in a remote rural setting), studies in certain settings have also previously reported superior survival associated with greater required travel to receive cancer care.1,2

Such data suggest the potentially meaningful impact of non– cancer-related factors, including the presence or absence of comorbidities (which might influence the ability to travel), the patient’s socioeconomic status (whether they have the resources to travel), and the availability of family support. The point is that although such factors may also affect clinical decision-making, their favorable or unfavorable influence on outcomes may be manifested regardless of the specific therapy selected.

A quite active debate within the gynecologic oncology arena relates to the issue in advanced ovarian cancer of the unique clinical utility associated with an attempt at maximal cytoreductive surgery at initial diagnosis vs an approach employing several cycles of neoadjuvant chemotherapy (after confirming a pathological diagnosis consistent with this malignancy), with a plan in most circumstances for interval cytoreduction in the absence of disease progression.3,4

To date, several published, peer-reviewed, phase 3 randomized trials in the advanced ovarian cancer setting conducted in different parts of the world have revealed that the two therapeutic strategies achieve similar survival outcomes, with the neoadjuvant approach being associated with a lower risk of serious postoperative morbidity and even mortality.3,4

These results have been challenged by other gynecologic cancer surgeons whose own reports (nonrandomized experiences) have suggested superior outcomes resulting from a primary surgical approach, with poorer outcomes observed in the patient population undergoing the neoadjuvant method. The argument is that the clinical judgment, skills, and experience, along with the know-how and resources available within their programs, permit a more aggressive, acceptably safe, and ultimately superior strategy.

However, it is precisely because of the role of effective clinical judgment in the ultimate selection of the patient population for the more aggressive, rather than less aggressive, approach, which appropriately assigns to the neoadjuvant strategy individuals with the most extensive (carcinomatosis) and difficult to manage circumstances (rapidly recurrent ascites or pleural effusions), as well as those with relevant serious comorbidities, that a direct comparison of survival outcomes is likely to be, at best, misleading.

To be clear, clinical judgment in determining the optimal choice of therapy is essential to enhancing the opportunity for the most favorable clinical outcome. But it is also critical to recognize that the ultimate survival outcome results from multiple factors, including inherent tumor biology as well as the response and potential resistance to the available antineoplastic drug therapy.

References

1. Lenhard Jr RE, Enterline JP, Crowley J, Ho GY. The effects of distance from primary treatment centers on survival among patients with multiple myeloma. J Clin Oncol. 1987;5(10):1640-1645. doi:10.1200/JCO.1987.5.10.1640
2. Lamont EB, Hayreh D, Pickett KE, et al. Is patient travel distance associated with survival on phase II clinical trials in oncology? J Natl Can Inst. 2003;95(18):1370-1375. doi:10.1093/jnci/djg035
3. Gaillard S, Lacchetti C, Armstrong DK, et al. Neoadjuvant chemotherapy for newly diagnosed, advanced ovarian cancer: ASCO guideline update. J Clin Oncol. 2025;43(7):868-891. doi:10.1200/JCO-24-02589
4. Bercow A, Stewart T, Bregar AJ, et al. Utilization of primary cytoreductive surgery for advanced-stage ovarian cancer. JAMA Netw Open. 2024;7(10):e2439893. doi:10.1001/ jamanetworkopen.2024.39893