Pre-admission polypharmacy burden and intensive care unit outcomes in patients with sepsis: A retrospective cohort study using the MIMIC-IV-ED linked database

This retrospective cohort study utilizing the MIMIC-IV-ED database demonstrates that high pre-admission polypharmacy (≥10 medications) serves as an independent predictor of increased ICU and 28-day mortality in sepsis patients, offering immediate prognostic value and highlighting specific high-risk drug interactions for targeted clinical intervention.

The Gist

Imagine your body as a busy, high-stakes control room. When a patient arrives at the hospital with sepsis (a severe, life-threatening reaction to an infection), it's like a massive fire alarm going off in that control room. The doctors rush in to put out the fire, but before they even start, they need to know what's already running on the control panel.

This study looked at a specific detail on that panel: how many different medicines the patient was already taking before they got sick.

Here is the breakdown of what the researchers found, using simple comparisons:

1. The "Medicine Load" Test

The researchers looked at over 3,300 adults who ended up in the Intensive Care Unit (ICU) for sepsis. They sorted these patients into three groups based on their "pre-admission medicine load":

• The Light Load: Taking 0 to 4 medicines.
• The Standard Load: Taking 5 to 9 medicines.
• The Heavy Load: Taking 10 or more medicines.

Think of this like a backpack. Some people were walking in with a light daypack, while others were carrying a massive hiking rucksack stuffed with gear. The study found that 59% of these patients were carrying that "Heavy Load."

2. The Connection to Survival

The researchers asked a simple question: Does carrying a heavier backpack make it harder to survive the fire?

The answer was a clear yes.

• Patients with no or few medicines had the best survival rates (about 18% did not survive).
• Those with a standard number of medicines had a higher risk (26% did not survive).
• Those with the "Heavy Load" (10+ medicines) had the highest risk (27.5% did not survive).

Even after the scientists adjusted for how sick the patients were when they arrived (making sure the "heavy backpack" wasn't just a sign of being sicker in general), the heavy load still stood out as a major risk factor. It's as if the backpack itself adds extra weight that makes climbing out of the danger zone much harder, regardless of how bad the fire is.

3. The "Frailty" of Too Many Pills

The paper calls this high number of medicines a form of "pharmacologic frailty."
Think of it like a car engine that is already running on too many different types of fuel additives. When a crisis hits, that engine is less flexible and more likely to stall. The study suggests that simply counting the pills gives doctors a quick, instant clue about how fragile a patient's system is before they even run complex tests.

4. Good Guys and Bad Guys in the Mix

The researchers also looked at which medicines were causing trouble:

• The Hero: Patients taking statins (often used for cholesterol) were actually less likely to die. It's like having a shield that helps the body fight the fire.
• The Villain Combo: Patients taking a specific mix of blood pressure drugs (RAS blockers) and water pills (diuretics) were much more likely to suffer kidney injury. It's like two tools clashing in the control room, causing a short circuit in the kidneys.

5. Why This Matters Right Now

The most exciting part of the study is how fast this information is available.

• When a patient walks into the Emergency Department, their list of current medicines is already written down.
• The study shows that a pharmacist or doctor can look at that list instantly (zero delay) and say, "This patient has a heavy medicine load; they are at higher risk."
• This isn't just a number; it helps identify specific dangerous combinations (like the blood pressure and water pill mix) that can be fixed immediately upon admission to the ICU.

In short: This study proves that if a sepsis patient arrives with a "heavy backpack" of 10+ daily medicines, they are in more danger than those with fewer. This information is free, instant, and helps doctors spot high-risk patients and dangerous drug clashes the moment they walk through the door.

Technical Summary

Technical Summary

Problem Statement
Polypharmacy is a highly prevalent condition among critically ill patients; however, its independent impact on intensive care unit (ICU) outcomes specifically within the context of sepsis remains critically unexplored. While the burden of multiple medications is recognized, there is a lack of evidence regarding whether pre-admission polypharmacy serves as an independent predictor of mortality or offers incremental prognostic value beyond standard severity scores.

Methodology
This study employed a retrospective cohort design utilizing the MIMIC-IV-ED linked database. The cohort consisted of 3,347 adults admitted to the ICU who met Sepsis-3 criteria. Pre-admission polypharmacy was stratified into three categories based on medication count: none (0–4 medications), standard (5–9 medications), and high (≥10 medications).

To evaluate outcomes, the researchers utilized multivariable logistic regression, propensity score matching, and reclassification analyses (Net Reclassification Improvement [NRI] and Integrated Discrimination Improvement [IDI]). The primary outcome was in-hospital ICU mortality, with secondary analysis of 28-day mortality and specific drug-class interactions. The study leveraged the medication reconciliation module available within the emergency department (ED) data to capture pre-admission medication loads.

Key Results

• Prevalence and Mortality Trends: High polypharmacy was present in 58.9% of the study population. A sequential increase in crude ICU mortality was observed across the groups: 18.5% for the "none" group, 26.0% for the "standard" group, and 27.5% for the "high" group (p < 0.001).
• Independent Prediction: After multivariable adjustment, high polypharmacy was found to independently predict in-hospital ICU mortality (adjusted Odds Ratio [aOR] 1.45; 95% CI 1.10–1.91) and 28-day mortality (aOR 1.47).
• Robustness: The association between high polypharmacy and mortality was confirmed through propensity score matching, which yielded a matched aOR of 1.28.
• Drug-Class Specific Findings:
  • Protective Effect: Statin use was identified as significantly protective (aOR 0.56).
  • Risk Interaction: The combination of Renin-Angiotensin System (RAS) blockers with diuretics was associated with an increased risk of acute kidney injury (aOR 1.49).
• Prognostic Value: The addition of polypharmacy status to the SOFA score provided significant early prognostic value, demonstrated by an NRI of 0.24.

Significance and Claims
The paper claims that high pre-admission polypharmacy represents a distinct form of "pharmacologic frailty" that is independent of acute illness severity. By utilizing the medication reconciliation table available in the ED, this metric offers a "zero-latency" tool for risk stratification at triage. The study concludes that this approach not only provides early prognostic value but also identifies specific, actionable high-risk drug interactions (such as RAS blockers plus diuretics) that can be targeted for immediate, pharmacist-led intervention upon ICU admission.