Eligibility criteria

Background

HemOnc.org has grown organically over time. From the beginning, the goal of the website has been to capture treatment details and the supporting literature for the systemic antineoplastic treatments used in the daily practice of hematology/oncology. As such, we initially drew our source material primarily from clinical practice guidelines, in particular focusing on treatments with a "top-level" recommendation (e.g., category 1 evidence in the NCCN Guidelines). As we have increasingly observed the standard of care evolve over time (e.g., immunotherapy was still in its infancy when the website was launched in 2011), we have developed an increasingly systematic approach to the identification and inclusion of regimens and their supporting literature. This is reflected in the first two of our four overarching goals:

• Create a database of all approved chemotherapy agents and supportive medications used in the field of hematology/oncology.
• Create a database of all standard-of-care chemotherapy regimens and references to primary literature (PubMed and direct links to the abstracts/full articles).

In the following sections, we describe our current approach (as of August 2019) to curating and selecting literature for inclusion on HemOnc.org.

Defining the Standard of Care (SOC)

One of the main goals of HemOnc.org is creating a database of all standard of care chemotherapy regimens. This is difficult because there is no universally accepted definition of standard of care beyond the legal definition. From Wikipedia, the legal definition of the medical standard of care is: "the level at which an ordinary, prudent professional with the same training and experience in good standing in a same or similar community would practice under the same or similar circumstances." We currently employ four separate definitions that meet the threshold of standard of care:

1. The control arm of a phase III randomized controlled trial (RCT). By implication, this means that all phase III RCTs with a control arm must eventually be included on the website. See prioritization below for more details.
2. The experimental arm(s) of a phase III RCT that provide(s) reasonable evidence (P-value less than 0.10) of superior efficacy for an intermediate surrogate endpoint (e.g., PFS) or a strong endpoint (e.g., OS)
3. A non-randomized study that is either:
   1. The basis for a regulatory agency approval (e.g., the US Food and Drug Administration [FDA])
   2. Recommended as a top-level regimen by ASCO, ESMO, or the NCCN
4. Any study (including case series and retrospective studies) that is specifically recommended by a member of our Editorial Board. All members of the Editorial Board with direct oversight of disease-specific pages are board-certified physicians (the "ordinary, prudent professional" standard).

Randomized trials

RCTs that have two control arms

These are generally non-inferiority trials. Regardless of the outcome, based on criteria #1 above, the study reference will be added to both control arms. If the control arm regimens are missing, they will be built and added to the website.

RCTs with one control arm and one or more experimental arms

Control arm(s)

• Regardless of whether the trial is "positive" or "negative", the study reference will be added to the control arm regimen. If the control arm regimen is missing, it will be built and added to the website.

Experimental arm(s)

• These will be added if the experimental arm has a statistically superior primary outcome, defined as p-value less than or equal to 0.10 (see levels of evidence)
• These will be generally not be added if the experimental arm has a statistically nonsignificant primary outcome, defined as p-value greater than 0.10 (see levels of evidence)
  • Exception 1: if a secondary outcome is the basis of a regulatory agency approval, we will add the regimen
  • Exception 2: if a member of the Editorial Board requests that the experimental arm be added, we will add the regimen
• Note: If the experimental arm(s) include a regimen with at least one investigational drug, the study reference will be added on the experimental drug's page, not as a detailed regimen on a disease-specific page. If and when the drug is approved by any regulatory agency, the information will be transferred from the experimental drug's page to the relevant disease-specific page(s).

RCTs with two or more experimental arms

This situation is very uncommon with Phase III trials, but relatively common with Randomized Phase II trials; it is a major reason why we relatively de-prioritize Randomized Phase II trials (see below). If the study is "negative" (i.e., the primary outcome is statistically nonsignificant), the study reference is not added to HemOnc.org. If the study is "positive", the study reference is added to the regimen which had the superior finding. If the regimen is missing, it will be built and added to the website.

Non-randomized or retrospective studies

Regimens consisting solely of approved drugs

• There is no strict cutoff for efficacy
• The regimen should be described by the authors as equivalent or better than standard-of-care, using such terms as "promising", "new standard-of-care", "at least as good", etc.
• If comparative efficacy to historic controls is reported, the new regimen should have a statistically superior outcome, defined as p-value less than or equal to 0.10 (see levels of evidence)
• The study should have at least 20 participants

Regimens that include an experimental drug

Note that if such a regimen is deemed eligible for inclusion, it will be added as a reference on the experimental drug's page, not as a detailed regimen on a disease-specific page. If and when the drug is approved, the information is transferred from the experimental drug's page to the relevant disease-specific page(s).

• The study should meet its predetermined primary efficacy endpoint
• It should be clear that the regimen will go on to be tested in phase III trials
  • Exception: if the regimen is the basis of an FDA approval, this is not required
• With few exceptions, the study should have at least 20 participants
  • If a phase I or Ib study, this should be at least 20 participants receiving the MTD

Prioritization

Because HemOnc.org is a voluntary effort driven by the contributors, it is not possible to add every study describing every treatment regimen, in anything close to real time. What follows is a general prioritization list for adding new regimens to the more heavily studied disease pages. In some diseases, there is no randomized literature so this prioritization does not apply:

1. Regimens described in registration studies (FDA has priority; other approval agencies are considered) whether they are randomized or not
2. Clinically relevant regimens, evaluated in fully enrolling phase III RCTs (including trials not specified as phase III but having a statistical power of 90% or greater), published in top-tier journals (New England Journal of Medicine, Lancet, JAMA, Journal of Clinical Oncology, Blood, Lancet Oncology, JAMA Oncology)
   1. Control arms of RCTs cited by the paper describing these regimens
   2. Experimental arms of RCTs cited by the paper describing these regimens, if they have statistically superior findings
3. Clinically relevant regimens, evaluated in fully enrolling phase III RCTs (including trials not specified as phase III but having a statistical power of 90% or greater), published in other journals (see Sources)
   1. Control arms of RCTs cited by the paper describing these regimens
   2. Experimental arms of RCTs cited by the paper describing these regimens, if they have statistically superior findings
4. Clinically relevant regimens, evaluated in randomized phase II RCTs or incompletely enrolled phase III RCTs (e.g., those closed early due to poor accrual), published in top-tier journals (New England Journal of Medicine, Lancet, JAMA, Journal of Clinical Oncology, Blood, Lancet Oncology, JAMA Oncology)
   1. Control arms of RCTs cited by the paper describing these regimens
   2. Experimental arms of RCTs cited by the paper describing these regimens, if they have statistically superior findings
5. Clinically relevant regimens, evaluated in randomized phase II RCTs or incompletely enrolled phase III RCTs (e.g., those closed early due to poor accrual), published in other journals (see Sources)
6. Clinically relevant regimens, evaluated in non-randomized trials, published in top-tier journals (New England Journal of Medicine, Lancet, JAMA, Journal of Clinical Oncology, Blood, Lancet Oncology, JAMA Oncology)
   1. Control arms of RCTs cited by the paper describing these regimens
   2. Experimental arms of RCTs cited by the paper describing these regimens, if they have statistically superior findings
7. Clinically relevant regimens, evaluated in non-randomized trials, published in other journals (see Sources)
8. Clinically relevant regimens, evaluated in fully enrolling phase III RCTs (including trials not specified as phase III but having a statistical power of 90% or greater), published in conference proceedings (see Sources)
9. Historically relevant regimens evaluated in RCTs, identified by other means not described above
10. Historically relevant regimens evaluated in non-randomized trials, identified by other means not described above

Disclaimer

Note: these eligibility criteria, which pertain mainly to treatment regimens, are guidelines that we try to follow rigorously. However, there are exceptions. For example, if a contributor or a member of our Editorial Board requests that a certain regimen be added, we will strive to do this even if the criteria are not strictly met. See the content tutorial and sources pages for more information. As outlined in our disclaimer, inclusion of a regimen on HemOnc.org is not an endorsement of its efficacy or appropriateness for use in any given clinical situation.