Favorable Implementation / QI

Oral / subcutaneous vs intravenous opioids

Opioid standard favoring oral and subcutaneous routes vs. Usual care (intravenous-preferred) · for Inpatient pain requiring opioids · real-time analysis of 5 studies · updated 2026-05-29

Inpatient stewardship programs that prioritize oral and subcutaneous opioids over intravenous routes cut IV opioid exposure by roughly 70–85% without worsening pain, and observational data link intravenous administration to substantially more adverse events and downstream opioid use.

Implementation / QI: Quality-improvement evidence with process/exposure outcomes (often pre-post, single-arm). Effects are rate ratios; a number-needed-to-treat generally does not apply. Glossary →

Interpretation & tips

Read these as exposure comparisons, not clinical-event trials. The pre/post programs (Ackerman, Amell, Dalal) cut IV opioid use 70–85% without worsening pain. NNT mostly does not apply—the outcomes are rates and adjusted odds. The adverse-event and downstream-opioid rows are observational, inverted to favor the oral/SC route. No randomized trial shows the strategy improves hard clinical outcomes.

80% lower risk of iv opioid exposure

Pooled RR 0.20 (95% CI 0.12–0.35) across 2 studies, 1,369 patients (random-effects, I² 92%).

0.20

Pooled RR

0.12–0.35

95% CI

Studies

1,369

Patients

Forest plot—IV opioid exposure

Study results—IV opioid exposure

Study	Design	Dose / regimen	Treatment	Control	RR [95% CI]	Improvement	Weight
Amell 2026 note	pre-post	Oral/SC-first opioid standard	—	—	0.27 [0.21–0.35]	73%	48%
Dalal 2021 excl note	pre-post	Multimodal nonopioid analgesia	32/104	105/241	0.71 [0.51–0.97]	29%	—
Ackerman 2018 note	pre-post	Oral/SC-first opioid standard	—	—	0.15 [0.13–0.18]	85%	52%

RR < 1 favors treatment for outcomes where lower is better. Rows in gray have a confidence interval crossing 1 (individually inconclusive). “~” marks effects reported as OR/HR and treated as RR-approximations. “excl” = excluded from pooling (e.g. reviews).

Notes & interpretation

Amell 2026—Rural replication of the Ackerman standard. IV opioid doses 0.158 → 0.043 per patient-day (73% reduction, p<0.001). Rate-ratio CI from the reported arm means and their 95% CIs. Pain scores did not increase (balancing measure).
Dalal 2021—Proportion of patients receiving ANY IV opioid: 43.6% → 30.8% (p=0.03). A different exposure metric than the dose-rate studies, so shown for context and excluded from the pooled IV-exposure estimate. Pain unchanged (3.9 → 3.7); LOS and 30-day readmission also fell.
Ackerman 2018—Original urban study. IV opioid doses 0.39 → 0.06 per patient-day (84% reduction, p<0.001). No CI was published; rate-ratio CI computed by treating doses as Poisson counts over patient-days (4,500 pre, 2,459 post). Pain did not worsen and improved by day 4-5.

40% lower risk of total opioid exposure (mme)

Pooled RR 0.60 (95% CI 0.46–0.78) across 2 studies, 1,024 patients (random-effects, I² 89%).

0.60

Pooled RR

0.46–0.78

95% CI

Studies

1,024

Patients

Forest plot—Total opioid exposure (MME)

Study results—Total opioid exposure (MME)

Study	Design	Dose / regimen	Treatment	Control	RR [95% CI]	Improvement	Weight
Amell 2026 note	pre-post	Oral/SC-first opioid standard	—	—	0.53 [0.48–0.58]	47%	52%
Ackerman 2018 note	pre-post	Oral/SC-first opioid standard	—	—	0.69 [0.60–0.80]	31%	48%

Notes & interpretation

Amell 2026—Total opioid MME 14.46 → 7.65 per patient-day (47% reduction, p<0.001). Rate-ratio CI from the reported arm means and their 95% CIs.
Ackerman 2018—Overall opioid MME 9.11 → 6.30 per patient-day (31% reduction). CI approximated from the reported means and SDs (n=patients); patient-day clustering not modeled, so the true interval is wider.

81% lower risk of opioid adverse events (oral/sc vs iv)

Pooled RR 0.19 (95% CI 0.15–0.24) across 1 studies, 31,742 patients (random-effects, I² 0%).

0.19

Pooled RR

0.15–0.24

95% CI

Studies

31,742

Patients

Forest plot—Opioid adverse events (oral/SC vs IV)

Study results—Opioid adverse events (oral/SC vs IV)

Study	Design	Dose / regimen	Treatment	Control	RR [95% CI]	Improvement	Weight
Daoust 2015 note	retrospective	Oral vs IV route	—	—	0.19 [0.15–0.24] ~	81%	100%

Notes & interpretation

Daoust 2015—Global adverse events, oral vs IV (inverted from the published adjusted IV-vs-oral OR 5.18, 4.13-6.49). Major adverse events oral vs IV ≈ 0.16 (inverted from 6.10). Subcutaneous is intermediate (SC-vs-oral OR 2.23).

70% lower risk of downstream opioid exposure

Pooled RR 0.30 (95% CI 0.16–0.59) across 1 studies, 862 patients (random-effects, I² 0%).

0.30

Pooled RR

0.16–0.59

95% CI

Studies

862

Patients

Forest plot—Downstream opioid exposure

Study results—Downstream opioid exposure

Study	Design	Dose / regimen	Treatment	Control	RR [95% CI]	Improvement	Weight
Dalal 2020 note	retrospective	Avoiding IV vs IV exposure	—	—	0.30 [0.16–0.59] ~	70%	100%

Notes & interpretation

Dalal 2020—Post-discharge opioid prescription, no-IV vs IV (inverted from the published adjusted IV-vs-no-IV OR 3.3, 1.7-6.4). In the IBD-flare subgroup the association was stronger (inverted OR ≈ 0.19).

Background

Intravenous opioids act faster and carry more euphoric potential than oral or subcutaneous routes, and observational data link inpatient IV exposure to more adverse events and later opioid use. Several hospitals adopted standards favoring oral, then subcutaneous, routes. The evidence is a mix of pre/post quality-improvement projects and observational cohorts.

Topic methodology & caveats

Rate ratios and inverted odds ratios; confidence intervals are derived from reported means or inverted from published IV-referenced estimates. Pre/post and observational designs limit causal inference.