Clinical article · Core

Point-of-care questions: what makes a clinical question answerable in under a minute

Written for
Hospital physicians, emergency doctors and residents
Reading time
7 minutes
Published
Updated

Written by the Sana Health team for clinicians. It summarises published guidance, is not a guideline and has not been peer reviewed. Check current local protocols before acting.

Key points

  • Clinicians raise many questions during patient care and are unable to pursue or answer most of them.
  • A foreground question about one decision for one patient can be answered quickly; a background question about a topic cannot.
  • A few patient details, such as renal function, pregnancy, weight, allergy and timing, change most answers and belong in the question.
  • A good quick answer restates the question, gives the action and its dose, names what should prompt escalation, and shows its evidence.
  • Short answers carry a risk of over-reliance; the clinician should check that the answer fits the patient and look at the sources.

Most questions go unanswered

Clinicians raise questions constantly while caring for patients. A systematic review of studies that observed them found that most of these questions are never pursued or never answered, and that unanswered questions may lead to poorer decisions [1]. An earlier review of primary care physicians found the same pattern: only a limited number of questions were followed up, usually by asking a colleague or checking a paper source first [2].

Lack of time is the usual reason, together with doubt that a usable answer exists. How the question is framed decides much of this. A precise question about one decision can often be answered from a guideline or a few trials in minutes; a broad one sends the clinician into a textbook.

Background and foreground questions

A background question asks about a topic: what causes hyponatraemia, how apixaban works. A foreground question asks about a decision for this patient: what to give, how much, how fast, or whether a test is needed. Foreground questions are the ones that can be answered at the point of care. The PICO structure (patient, intervention, comparison, outcome) is a reliable way to turn one into the other.

VagueAnswerable
Antibiotics for UTI?First-line oral treatment for uncomplicated cystitis in a non-pregnant woman with eGFR 38?
How do you reverse apixaban?Reversal options for apixaban, last dose 4 hours ago, in a patient with intracerebral haemorrhage?
Low sodium management?Maximum 24-hour rise for chronic hyponatraemia of 112 mmol/L in a malnourished patient who drinks heavily?

The details that change the answer

Most quick questions fail because a single detail is missing, not because the evidence is thin. The following change the answer often enough that they belong in the question whenever they apply.

  • Age, and weight when a dose depends on it.
  • Renal function, with the measure used (eGFR or creatinine clearance), and hepatic function.
  • Pregnancy or breastfeeding.
  • Allergies, and what the reaction actually was.
  • Current medicines that interact, including anticoagulants and antiplatelets.
  • Timing: when symptoms began, or when the last dose was taken.
  • Severity and stability: whether the patient is haemodynamically stable, symptomatic or deteriorating.
  • What is available locally, such as a reversal agent or a formulary choice.

Three worked examples

Cystitis with an eGFR of 38. The IDSA and ESCMID guideline defines uncomplicated cystitis in premenopausal, non-pregnant women without known urological abnormalities, and recommends first-line agents including nitrofurantoin, trimethoprim-sulfamethoxazole where resistance is low, fosfomycin and pivmecillinam [4]; short courses are appropriate [5]. The eGFR matters because renal thresholds for nitrofurantoin differ between the US label, national formularies and prescribing criteria for older adults, and 38 falls in the range where they disagree. A good answer says which source it applied and defers to local guidance.

Apixaban in life-threatening bleeding. The choice depends on the site and severity of bleeding, the time of the last dose, renal function and what the hospital stocks. Reversal strategies are shaped by the pharmacology of the anticoagulant and antidote, regulatory indications, the quality of evidence and availability [6], and guidance from the ISTH and the European Society of Anaesthesiology and Intensive Care addresses the options [7, 8]. In a randomised trial in factor Xa inhibitor-associated intracerebral haemorrhage, andexanet improved haemostatic efficacy compared with usual care but was associated with more thrombotic events [9].

Chronic hyponatraemia of 112 mmol/L. The European guideline limits the rise in sodium during the first 24 hours, and the US and European guidance differ in detail [10, 11]. The answer depends on whether the patient has severe symptoms, how long the hyponatraemia has been present, and whether there are risk factors for osmotic demyelination such as alcohol use, malnutrition, hypokalaemia or liver disease. A question that states them can be answered precisely; one that does not gets a range.

A template for the question

  1. State the decision: what to give, what dose, how fast, whether to test or refer.
  2. Describe the patient in one line, including only the details that change the decision.
  3. Add what has already been done and what is available.
  4. Say what matters most: safety in renal impairment, avoiding a drug class, speed of effect.

If the patient is unstable, the diagnosis is uncertain, or the findings do not fit together, the problem is not a quick question. It needs a full assessment and senior input, not a shorter answer.

Reading a short answer safely

Decision support brings its own errors. A systematic review of automation bias found that users of clinical decision support tend to over-rely on its output and may fail to notice the new errors it introduces [3]. A short answer should therefore be read actively: does it restate the question correctly, does the dose fit the patient's weight and renal function, and do the cited sources say what the answer claims? A systematic review of how large language models are tested in health care concluded that current evaluation approaches may not identify where they are most useful [12], a further reason to check the evidence behind any quick answer before acting on it.

How Sana Health helps

Rapid Consult takes one specific question and the few patient details that change it, and returns a short report: a direct answer with a statement of confidence, up to five actions in order with doses written out, the findings that should prompt escalation, and the two to four records that best support the answer. The full reference list is read from PubMed and ClinicalTrials.gov, with US product information when a medicine is named, and the report typically arrives in under a minute. If the question cannot be answered safely in brief, it says so.

Questions

What is a PICO question?

A clinical question built from four parts: the patient or problem, the intervention being considered, the comparison, and the outcome that matters. It turns a general topic into a specific decision that the evidence can answer.

Why do clinicians leave so many questions unanswered?

Mainly lack of time, and doubt that a usable answer exists. Studies of clinicians at the point of care show that most questions raised are never pursued or answered.

What should be done when a quick answer does not fit the patient?

Treat the mismatch as a signal. Check that the question included the details that matter, read the sources cited, and take the case to a full assessment or a senior colleague if it is more complex than a single question.

References

  1. Clinical questions raised by clinicians at the point of care: a systematic review (opens in a new tab)JAMA Intern Med, 2014
  2. How do primary care physicians seek answers to clinical questions? A literature review (opens in a new tab)J Med Libr Assoc, 2006
  3. Automation bias: a systematic review of frequency, effect mediators, and mitigators (opens in a new tab)J Am Med Inform Assoc, 2012
  4. International clinical practice guidelines for the treatment of acute uncomplicated cystitis and pyelonephritis in women: A 2010 update by the Infectious Diseases Society of America and the European Society for Microbiology and Infectious Diseases (opens in a new tab)Clin Infect Dis, 2011
  5. Appropriate Use of Short-Course Antibiotics in Common Infections: Best Practice Advice From the American College of Physicians (opens in a new tab)Ann Intern Med, 2021
  6. Antidotes for Anticoagulation Reversal (opens in a new tab)N Engl J Med, 2026
  7. Reversal of direct oral anticoagulants: guidance from the SSC of the ISTH (opens in a new tab)J Thromb Haemost, 2024
  8. Clinical guideline on reversal of direct oral anticoagulants in patients with life threatening bleeding (opens in a new tab)Eur J Anaesthesiol, 2024
  9. Andexanet for Factor Xa Inhibitor-Associated Acute Intracerebral Hemorrhage (opens in a new tab)N Engl J Med, 2024
  10. Clinical practice guideline on diagnosis and treatment of hyponatraemia (opens in a new tab)Intensive Care Med, 2014
  11. Diagnosis and Treatment of Hyponatremia: Compilation of the Guidelines (opens in a new tab)J Am Soc Nephrol, 2017
  12. Testing and Evaluation of Health Care Applications of Large Language Models: A Systematic Review (opens in a new tab)JAMA, 2025