§ 01 · Scientific governance
Each is enforced in the release process, not left to good intentions.
01
Written rules, not generated answers.
Daleel Verify calculates with written rules, each shown with its source and version. No dose is produced by a model.
02
Supervised by clinical pharmacology.
The rules are developed under the supervision of specialists in clinical pharmacology and are subject to scientific review before clinical use.
03
Every active rule is tested.
No rule is released without a verification case that triggers it. A rule with no case stops the release.
04
Change is declared.
A release that changes any clinical result must name the case and the reason, or it does not ship.
05
Language never changes the verdict.
Arabic and English produce identical results, tested by automated comparison.
06
Evidence and preference kept apart.
An institution’s preference is shown beside the evidence, never inside it.
§ 02 · Editorial policy
How evidence enters the product, and how it is shown to clinicians.
01
Sources are declared.
Every rule names the evidence it rests on: a trial, a guideline or the prescribing information.
02
Guideline text is cited, never copied.
We cite the guideline, its version and section, and summarise in our own words until a licence is signed.
03
Excerpts wait for approval.
Prescribing information excerpts are shown to clinicians only after our clinical team approves each one.
04
Nothing is assumed.
An empty field is never filled with a default. Missing data is reported as missing.
§ 03 · Validation method
Agreement with past results is not clinical validation. Our protocol measures the engine against expected results written independently by clinicians, and reports performance for every rule with its uncertainty.
Lower bound of sensitivity at 95% confidence, when a rule detects every positive case
72%
10 cases
84%
20 cases
89%
30 cases
94%
60 cases
Wilson score interval. Used to plan the number of cases per rule.
§ 04 · Evidence library
A selection of the pivotal studies behind the breast cancer pathway, summarised in our own words.
§ 05 · The Saudi evidence
Saudi women are diagnosed younger. Records are often incomplete. Documentation decides approvals.
25,864[1]
new cancer cases recorded in the Kingdom in 2023.
5,207[1]
new breast cancer cases among women in 2023, the most common cancer in women.
17.3%[2]
of new cancers among Saudi nationals in 2022 were breast cancer, the leading cancer.
~50[3]
typical age at breast cancer diagnosis in Saudi women, against 60 to 65 in Western populations.
53%[4]
of records had no recorded height in a Saudi hospital study. Weight was missing in 47%, allergies in 52%.
47.3%[5]
of 203 rejected claims in two Saudi public hospitals were rejected for documentation.
§ 06 · Research collaboration
We work with academic affairs and research centres on joint, peer reviewed studies of medication safety in oncology, led by your clinicians.
The guidelines and references our clinical content is built on, with the edition each one follows.