Can I conduct a retrospective chart‑review study on the agreement between Microsoft Dragon Copilot recommendations and clinician disposition decisions for adult out‑patient urinary‑tract infections, and is it true that no such studies have been published in the Philippines?

Medical Advisory BoardAll articles are reviewed for accuracy by our Medical Advisory Board
Educational purpose only • Exercise caution as content is pending human review
Article Review Status
Submitted
Under Review
Approved

Last updated: October 2, 2026 • View editorial policy

Have a follow-up question?

Our Medical A.I. is used by practicing medical doctors at top research institutions around the world. Ask any follow up question and get world-class guideline-backed answers instantly.

Ask Question

Magandang umaga po! Hindi po ito mainam na pag-aaral na isagawa.

Ang inyong iminumungkahing retrospective chart review ay may malubhang metodolohikal na problema at hindi angkop para sa pagsusuri ng AI-driven clinical decision support system tulad ng Microsoft Dragon Copilot sa UTI management.

Mga Pangunahing Dahilan Bakit Hindi Ito Mainam

1. Mali ang Study Design para sa AI Clinical Decision Support Evaluation

Ang DECIDE-AI guideline 1, na inilathala sa Nature Medicine (2022), ay malinaw na nagsasaad na ang early-stage evaluation ng AI-driven decision support systems ay nangangailangan ng:

  • Prospective observational design (hindi retrospective)
  • Real-time documentation ng user exposure at adherence sa AI system
  • Systematic recording ng human-computer agreement patterns
  • Active monitoring ng safety events at errors habang nangyayari

Ang retrospective chart review ay hindi makakakuha ng mga kritikal na datos na ito dahil:

  • Hindi ninyo makikita kung paano talaga ginamit ng clinician ang AI tool sa real-time
  • Walang documentation kung bakit sumunod o hindi sumunod ang clinician sa AI recommendation
  • Hindi ninyo malalaman kung may technical errors o malfunctions na naganap
  • Walang usability evaluation data

2. Kulang ang Inyong Proposed Methodology

Ang DECIDE-AI checklist 1 ay nag-require ng:

  • Item 10a: User exposure metrics, number of AI system uses, adherence patterns
  • Item 12: Detailed human-computer agreement analysis, including reasons for variation
  • Item 13: Real-time safety monitoring at error documentation
  • Item 14: Usability evaluation using recognized frameworks

Ang chart review lang ay hindi sapat para makuha ang lahat ng ito.

3. May Mas Angkop na Study Design

Ang recent study sa Israel 2 (2026, Scientific Reports) ay gumamit ng prospective cohort design para sa similar na tanong tungkol sa AI-driven UTI decision support. Nakita nila na:

  • 33% ng physicians ay hindi sumunod sa AI recommendations
  • Kailangan ng real-time data collection para maintindihan ang adoption patterns
  • Physician characteristics (edad, practice setting, patient volume) ay nag-influence sa AI adoption

Kung retrospective chart review lang ang gagawin ninyo, hindi ninyo makikita ang mga patterns na ito.

4. Tungkol sa Claim na "No Studies in the Philippines"

Hindi ninyo pwedeng i-claim na "no studies conducted yet in the entire Philippines" based sa literature search lang. Ang absence of published evidence ay hindi proof of absence. Maaaring may:

  • Ongoing studies na hindi pa published
  • Institutional quality improvement projects
  • Thesis o dissertation work
  • Conference presentations na hindi pa naging full papers

Ang Dapat Ninyong Gawin Imbes

Recommended Alternative Approach:

  1. Prospective Observational Study (tulad ng sinuggest ng DECIDE-AI 1)

    • Real-time documentation ng AI tool usage
    • Structured data collection ng clinician decisions
    • Systematic recording ng agreement/disagreement patterns
    • Active safety monitoring

  2. Minimum Data Collection Requirements:

    • Baseline clinician characteristics
    • Each instance ng AI system use
    • Clinician adherence sa AI recommendations
    • Reasons for variation from AI suggestions
    • Patient outcomes (clinical failure, hospitalization) 3
    • Usability feedback from clinicians

  3. Follow CHEERS Guidelines kung may cost component 4

  4. Ethics Considerations:

    • Informed consent kung prospective
    • Data privacy protections
    • Algorithm fairness assessment

Mga Pitfalls na Iwasan

  • Huwag mag-rely sa incomplete chart documentation - ang retrospective charts ay kulang sa detalye tungkol sa clinical decision-making process
  • Huwag mag-assume na chart review ay "easier" - mas mababa ang quality ng data, mas mahirap ang valid conclusions
  • Huwag mag-claim ng "first in Philippines" nang walang comprehensive verification

Konklusyon ng Rekomendasyon

Kung gusto ninyong mag-conduct ng meaningful evaluation ng AI clinical decision support para sa UTI, kailangan ninyo ng prospective design na sumusunod sa DECIDE-AI framework 1. Ang retrospective chart review ay hindi sapat at maaaring magbigay ng misleading conclusions tungkol sa AI system performance at clinician adoption patterns.

Kung limited ang resources, mas mainam pang mag-focus sa smaller prospective pilot study kaysa malaking retrospective review na may fundamental methodological flaws.

References

3
Research

Outpatient treatment and clinical outcomes of bacteriuria in veterans: A retrospective cohort analysis.

Antimicrobial stewardship & healthcare epidemiology : ASHE, 2022

Related Questions

What is the proper dosage of uva ursi for treating urinary tract infections (UTIs)?
What is the recommended dose of antibiotics for urinary tract infection (UTI) treatment?
Is Ural (methenamine) effective for urinary tract infections (UTIs)?
What is the management of urinary tract infections (UTIs) according to the Infectious Diseases Society of America (IDSA) guidelines?
What is the best antibiotic regimen for a patient with pneumonia being treated with doxycycline (doxy) who has also been diagnosed with a urinary tract infection (UTI)?
Does eating a normal portion of fish, whether grilled, fried, or in a burger, raise blood pressure?
What is the appropriate workup and management for a patient presenting with sicca syndrome and periorbital edema?
What is the differential diagnosis for a concussion?
Does consumption of dried fruits increase blood pressure?
When an adult patient leaves the hospital against the medical staff's wishes, must the physician remain present?
Do unsalted, unsweetened dried fruits such as peanuts, almonds, cashews, pistachios, and raisins increase blood pressure?

Professional Medical Disclaimer

This information is intended for healthcare professionals. Any medical decision-making should rely on clinical judgment and independently verified information. The content provided herein does not replace professional discretion and should be considered supplementary to established clinical guidelines. Healthcare providers should verify all information against primary literature and current practice standards before application in patient care. Dr.Oracle assumes no liability for clinical decisions based on this content.