Patient Care

Reviews of Research on Haloperidol and ICU Delirium, Proton Pump Inhibitors, Thrombolytics and Stroke


In This Edition

Literature At A Glance

A guide to this month’s studies

  1. Intravenous haloperidol does not prevent ICU delirium
  2. Predicting delirium risk in hospitalized adults
  3. Oral PPIs as effective as IV PPIs in peptic ulcer bleeding
  4. Probiotic benefit questioned in the elderly
  5. Colchicine and NSAID better than NSAID alone for acute pericarditis
  6. Improvement needed in patient understanding at hospital discharge
  7. Effectiveness of a multihospital effort to reduce rehospitalization
  8. Hospitals profit from preventing surgical site infections
  9. Prothrombin complex concentrate safer than fresh frozen plasma in rapidly reversing INR
  10. Hospital-acquired anemia associated with higher mortality, increased LOS
  11. Thrombolytics and stroke: the faster the better

Intravenous Haloperidol Does Not Prevent ICU Delirium

Clinical question: Can haloperidol reduce delirium in critically ill patients if initiated early in ICU stay?

Background: Prior studies suggest antipsychotics reduce intensity and duration of delirium in hospitalized patients. Evidence is mixed for preventing delirium. A trial of risperidone demonstrated delirium rate reduction in coronary artery bypass grafting (CABG) patients, but another trial of haloperidol in hip surgery patients failed to prevent onset of delirium. There is little evidence on antipsychotics in ICU delirium.

Study design: Randomized, double-blinded, placebo-controlled trial.

Setting: Single, adult ICU in England.

Synopsis: The study randomized 142 critically ill patients to receive 2.5 mg of intravenous haloperidol versus placebo every eight hours for up to 14 days. There was no significant difference between groups in the total time spent free of delirium or coma. Limitations include the use of open-label haloperidol in 21% of the placebo group patients. More sedation but less agitation was seen with the use of haloperidol, which also prolonged QTc. No severe adverse effects were observed.

This study supports the idea that scheduled antipsychotics should not be used to reduce ICU delirium. Addressing modifiable risk factors and using dexmedetomidine rather than lorazepam for sedation in the ICU continue to be first-line strategies to lower delirium rates.

Bottom line: Prophylactic haloperidol should not be used to prevent ICU delirium.

Citation: Page VJ, Ely EW, Gates S, et al. Effect of intravenous haloperidol on the duration of delirium and coma in critically ill patients (Hope-ICU): a randomized, double-blind, placebo-controlled trial. Lancet Respir Med. 2013;1(7):515-523.

Predicting Delirium Risk in Hospitalized Adults

Clinical question: Can a simple tool be developed and used for predicting delirium in hospitalized adults?

Background: Delirium is a common condition that results in higher mortality, longer length of stays, and higher probability of discharge to nursing home. Current delirium prediction tools are complicated, or restricted to surgical or critically ill patients.

Study design: Prospective cohort study, with separate derivation and validation cohorts.

Setting: Two academic hospitals and a VA hospital in San Francisco.

Synopsis: Investigators enrolled 374 hospitalized patients who were more than 50 years of age and not delirious at time of admission (209 patients in the derivation and 165 in the validation). The authors identified four predictors of delirium: Age >80; failure to spell “World” backwards; disOrientation to place; and higher nurse-rated iLlness severity (AWOL). The authors found that rates of delirium increased with increasing number of predictors (with zero predictors, 2% developed delirium; one predictor, 4%; two predictors, 14%; three predictors, 20%; four predictors, 64%).

These predictors are similar to other previously identified risk factors, as well as to prediction tools that are in use for surgical patients. However, this tool is quick and can be completed by nursing staff, so it may have a role to play in helping triage patients to units more specialized in preventing delirium.

Bottom line: The AWOL prediction tool is simple to use, broadly applicable, and adds another tool to the literature to determine delirium risk.

Citation: Douglas VC, Hessler CS, Dhaliwal G, et al. The AWOL tool: derivation and validation of a delirium prediction rule. J Hosp Med. 2013;8(9);493-499.

Oral Proton Pump Inhibitors (PPIs) as Effective as IV PPIs in Peptic Ulcer Bleeding

Clinical question: In patients with peptic ulcer bleeding, are oral PPIs of equal benefit to intravenous PPIs?

Background: PPI therapy has been shown in several studies to reduce re-bleeding risk in patients when used adjunctively for peptic ulcer bleeding. In spite of this data, there is still uncertainty about the optimal dose and route of administration.

Study design: Meta-analysis of prospective, randomized control trials.

Setting: OVID database search in June 2012.

Synopsis: A literature search identified six prospective, randomized control trials. Overall, 615 patients were included across the six trials. No significant difference in risk of re-bleeding was discovered between the two groups (8.6% oral vs. 9.3% IV, RR: 0.92, 95% CI: 0.56-1.5). Length of hospital stay was statistically significantly lower for oral PPIs (-0.74 day, 95% CI: -1.10 to -0.39 day).

Because these findings are based on a meta-analysis of studies with notable flaws—including lack of blinding—it is difficult to draw any definitive conclusions from this data. Hospitalists should use care before changing their practice patterns, given the risk of bias and need for further study.

Bottom line: Oral PPIs may reduce hospital length of stay without an increased risk of re-bleeding; however, further study with a well-powered, double-blind, randomized control trial is necessary.

Citation: Tsoi KK, Hirai HW, Sung JJ. Meta-analysis: comparison of oral vs. intravenous proton pump inhibitors in patients with peptic ulcer bleeding. Aliment Pharmacol Ther. 2013;38(7):721-728.

Probiotic Benefit Questioned in the Elderly

Clinical question: Do probiotics prevent antibiotic-associated diarrhea (AAD) in patients 65 and older?

Background: Individual studies using different protocols to assess the efficacy of probiotics in preventing AAD, including Clostridium difficile-associated diarrhea (CDAD), suggest a decreased incidence of AAD when taking probiotics. Meta-analysis of this data also suggests that probiotics are effective in prevention of AAD; however, these results are undermined by the high heterogeneity of the studies included.

Study Design: Randomized, double-blind, placebo-controlled trial.

Setting: Multicenter trial in the United Kingdom.

Synopsis: Nearly 3,000 patients ages 65 years and older who had received one or more antibiotics within seven days were randomized to receive placebo or high-dose probiotics for 21 days. After recruitment, the patients were assessed for AAD up to eight weeks and CDAD up to 12 weeks. Results did not demonstrate a reduction of AAD or CDAD in patients taking probiotics. AAD occurred in 10.8% of patients taking the probiotic and 10.4% of patients taking placebo (95% confidence interval 0.83-1.32). CDAD occurred in 0.8% of patients taking the probiotic and 1.2% of patients taking placebo (95% confidence interval 0.34-1.47).

Based on the results of this double-blind, placebo-controlled trial, there is insufficient evidence to support initiation of probiotics for the prevention of AAD and CDAD in patients 65 years and older. Future studies utilizing standardized protocols against specific antibiotics, along with improved understanding of the underlying mechanisms of AAD prevention, are needed.

Bottom line: High-dose probiotics (lactobacillus acidophilus and bifidobacterium bifidum) do not prevent AAD in elderly patients.

Citation: Allen S, Wareham K, Wang D, et al. Lactobacilli and bifidobacteria in the prevention of antibiotic-associated diarrhoea and Clostridium difficile diarrhoea in older inpatients (PLACIDE): a randomised, double-blind, placebo-controlled, multicentre trial. Lancet. 2013;382(9900):1249-1257.

Colchicine and NSAID Better than NSAID Alone for Acute Pericarditis

Clinical question: Is colchicine safe, effective, and able to prevent recurrence in acute pericarditis?

Background: Colchicine is effective for the treatment of recurrent pericarditis. More recent open-label trials have established its role in acute pericarditis when combined with conventional NSAID therapy. However, a definitive randomized control trial has not been performed to establish colchicine’s role in acute pericarditis.

Study design: Double-blinded, randomized, controlled trial.

Setting: Multicenter in Northern Italy.

Synopsis: Investigators randomized 240 patients to receive either colchicine or placebo in addition to standard therapy of either aspirin or ibuprofen. Incessant or recurrent pericarditis occurred in 16.7% of patients treated with colchicine versus 37.5% in patients receiving placebo (RR 0.56; 95% CI 0.30-0.72; P<0.001). The number needed to treat to prevent one episode of incessant or recurrent pericarditis was four. Colchicine therapy also reduced the frequency of symptom persistence at 72 hours, number of recurrences per patient, rate of hospitalization, and the rate of readmission within one week.

It should be noted that the study excluded the following groups: patients with an elevated troponin, elevated transaminases (>1.5 upper limit of normal), and serum creatinine >2.5.

Bottom line: In addition to conventional therapy, colchicine reduces incessant or recurrent pericarditis in patients with a first episode of acute pericarditis.

Citation: Imazio M, Brucato A, Cemin R, et al. A randomized trial of colchicine for acute pericarditis. N Engl J Med. 2013;369(16):1522-1528.

Although daily charges were essentially the same between the groups, patients with SSIs had almost double the mean length of stay than patients without SSIs. SSI patients also had a drastically higher 30-day readmission rate.

Improvement Needed in Patient Understanding at Hospital Discharge

Clinical question: How well do older patients with heart failure, pneumonia, or acute coronary syndrome understand their diagnosis and post-discharge follow-up plans compared with medical record data?

Background: As hospitals across the country work on preventing 30-day readmissions, more attention has been given to assessing the quality of discharge processes; few evaluations have been conducted from a patient-centered perspective.

Study Design: Prospective, observational cohort study.

Setting: An urban, academic medical center.

Synopsis: This study evaluated the quality of the discharge process among 377 hospitalized patients >65 years old. Medical record data were compared with patient responses during a telephone interview within one week of discharge. By medical records, every patient received discharge instructions that in 97% of cases included discharge diagnosis, activity instructions, follow-up physician information, and warning signs. The authors determined that discharge diagnosis was not written in lay terms 26% of the time. By patient report, 90% expressed that they understood their discharge diagnosis, yet only around 60% fully understood their diagnosis as it was written in the medical record. Although about half of patients reported having a follow-up appointment upon discharge, only about a third of patients had a documented follow-up appointment in the medical record.

Bottom line: Multiple discrepancies were identified between medical record review and patients’ understanding of their discharge diagnosis and plans. Improvements in discharge processes (such as making follow-up appointments) and in patient education (such as increased use of layperson language) are needed.

Citation: Horwitz LI, Moriarty JP, Chen C, et al. Quality of discharge practices and patient understanding at an academic medical center [published online ahead of print August 19, 2013]. JAMA Intern Med.

Effectiveness of a Multihospital Effort to Reduce Rehospitalization

Clinical question: Does Project BOOST reduce 30-day rehospitalization for hospitals participating in a quality improvement collaborative?

Background: With the advent of penalties for hospitals with excessive 30-day readmissions among Medicare beneficiaries, hospitals nationwide are attempting to reduce 30-day readmission rates. Few interventions aimed at reducing 30-day hospital readmissions have been effective, and successful interventions have limited generalizability.

Study design: Semi-controlled, pre-post study.

Setting: Volunteer sample of acute care pilot units within a nationally representative sample of 11 academic and non-academic hospitals.

Synopsis: The 11 hospitals enrolled in this quality improvement collaborative planned and implemented Project BOOST tools over 12 months with support from an external quality improvement mentor. Each hospital tailored the BOOST tools that they implemented based on a needs assessment. Reporting of clinical outcome data was voluntary; administrative sources at each hospital provided these data. Although 30 hospitals participated in this collaborative, only 11 hospitals reported data for this analysis.

Average 30-day rehospitalization rates among BOOST units fell from pre- to post-implementation (14.7% to 12.7%, P=0.010); 30-day rehospitalization rates among control units did not change during this same time period (14.1% to 14.0%, respectively, P=0.831).

Bottom line: Although the 11 hospitals in this collaborative found reduced 30-day readmissions in association with BOOST implementation, this finding may be biased due to voluntary reporting of data and improvements at one hospital driving the overall effect of the intervention. More rigorous evaluation of Project BOOST is needed.

Citation: Hansen LO, Greewald JL, Budnitz T, et al. Project BOOST: Effectiveness of a multihospital effort to reduce rehospitalization. J Hosp Med. 2013;8(8):421-427.

Hospitals Profit from Preventing Surgical Site Infections

Clinical question: Does quality improvement, in this case preventing surgical site infections (SSIs), necessarily lead to improvement in hospital profit?

Background: It’s clear that preventing SSIs benefits patients and saves money for health insurance providers, but it’s unclear what financial impact SSIs have on hospitals and how best to calculate it. This quantification is needed for cost-benefit analyses of interventions designed to prevent SSIs.

Study design: Retrospective study.

Setting: Four Johns Hopkins-affiliated, tertiary care hospitals.

Synopsis: This retrospective study included all patients admitted to or having certain surgical procedures at four Johns Hopkins-affiliated hospitals between Jan. 1, 2007, and Dec. 31, 2010. Patients were first stratified by complexity, and then those who had a SSI (618) were compared to those without SSIs (399,627 admissions and 25,849 surgeries) for differences in daily hospital charges, length of stay, 30-day readmission rates, and hospital profit.

Although daily charges were essentially the same between the groups, patients with SSIs had almost double the mean length of stay than patients without SSIs. SSI patients also had a drastically higher 30-day readmission rate.

The authors propose equations to determine the change in hospital profit due to a single SSI and calculated that preventing one SSI led to an increase in hospital profit between $4,147 and $22,239. These numbers haven’t included the cost of a SSI prevention program, and the limitations in applying these numbers to all hospitals include widely varying hospital costs and differing ability to fill empty beds.

Bottom line: In these four tertiary care hospitals, each SSI prevented could increase hospital profit by thousands of dollars, as well as significantly decrease length of stay and 30-day readmission rates.

Citation: Shepard J, Ward W, Milstone A, et al. Financial impact of surgical site infections on hospitals: the hospital management perspective. JAMA Surg. 2013;148(10):907-914.

Prothrombin Complex Concentrate Is Safer than Fresh Frozen Plasma in Rapidly Reversing INR

Clinical question: Is prothrombin complex concentrate (PCC) safer and more effective than fresh frozen plasma (FFP) in reversing international normalized ratio (INR)?

Background: In Canada, PCC has become the standard of care over FFP for reversal of critical INR due to decreased time of administration, faster preparation, lack of allergic reactions, and small volume. Few studies compare these two products in their adverse effects, time to INR reversal, length of stay, and blood transfusion requirements.

Study design: Retrospective cohort study.

Setting: Two tertiary care EDs in Canada.

Synopsis: Health records of adult patients with an INR ≥1.8 who received FFP over a two-year period prior to PCC introduction (n=149) were compared to those who received PCC in the two years after PCC introduction (n=165). Total serious adverse events, which include mortality, myocardial infarction, and heart failure, were higher in the FFP group (19.5% versus 9.7%, P=0.0164). Heart failure exacerbations, time to reversal of INR, and units of blood transfused were increased in the FFP group. There was no difference in thromboembolic events or in length of stay.

Due to this study’s retrospective nature, there were issues with documentation of INR measurements, so true rapidity of INR reversal is unknown. In the United States, the FDA only recently approved PCC for use, so availability might be limited.

Bottom line: Prothrombin complex concentrate is an effective and fast alternative to FFP for reversal of critical INR levels.

Citation: Hickey M, Gatien M, Taljaard M, Aujnarain A, Giulivi A, Perry JJ. Outcomes of urgent warfarin reversal with frozen plasma versus prothombin complex concentrate in the emergency department. Circulation. 2013;128(4):360-364.

Hospital-Acquired Anemia Associated with Higher Mortality, Increased LOS

Clinical question: What is the prevalence of hospital-acquired anemia (HAA), and does it lead to increased mortality and resource utilization?

Background: HAA is a multifactorial care-based problem that occurs as a result of hemodilution, phlebotomy, blood loss from procedures, and impaired erythropoiesis. In the general hospital population, very little is known about HAA prevalence or whether HAA is associated with increased mortality, greater length of stay (LOS), or higher costs.

Study design: Retrospective cohort study.

Setting: Large academic health system in Ohio.

Synopsis: Using administrative data and electronic health record data, an analysis of 188,447 hospitalizations showed that HAA prevalence was 74%. Worsening HAA was correlated to an increase in mortality, so that the odds ratio of mortality with moderate anemia (Hgb between >9 and ≤11) was 1.51 (95% confidence interval 1.33-1.71, P<0.001) and severe anemia (Hgb ≤9) was 3.28 (95% confidence interval 2.90-3.72, P<0.001). Increased degree of HAA was correlated to increasing LOS (up to 1.88 extra days for patients with severe anemia) and higher hospital costs.

Because this is a retrospective observational study, no true causal relationship can be discerned from this study. However, the body of evidence linking iatrogenic causes of anemia to negative outcomes is compelling. Hospitalists should attempt to limit blood loss through judicious use of phlebotomy and procedures in their patients, so as to avoid anemia and subsequent unnecessary transfusions.

Bottom line: Hospital-acquired anemia is associated with higher mortality, LOS, and hospital costs in all hospitalized patients.

Citation: Koch CG, Li L, Sun Z, et al. Hospital-acquired anemia: prevalence, outcomes, and healthcare implications. J Hosp Med. 2013;8(9):506-512.

Thrombolytics and Stroke: The Faster the Better

Clinical question: Does time from ischemic stroke onset to treatment with intravenous thrombolysis make a difference?

Background: Previous studies have shown that “time is brain.” Quicker treatment with intravenous thrombolysis improves outcomes. Multicenter comparison of very early treatment (i.e., <90 minutes) to a later onset to treatment has not been done.

Study design: Observational study.

Setting: Patient information from 1998 to 2012 from 10 European stroke centers.

Synopsis: A total of 6,856 patients were included, of which 19% received thrombolysis in <90 minutes. None of the patients received endovascular treatment for stroke. Modified Rankin score, a functional assessment, was used to determine outcome. A score of 0 or 1, an “excellent” outcome, was seen more often in patients with a moderate severity stroke (NIH stroke scale of 7-12) who received thrombolysis in <90 minutes, but not in other groups. Thrombolysis in <90 minutes was associated with fewer intracerebral hemorrhages (ICH), but symptomatic ICH was not statistically significantly different. Mortality at three months was not different in the two time groups.

Limitations to this study included an unknown presumed cause of stroke in more than a quarter of patients. Deviations from acute stroke protocols are not described. This study adds to the body of literature supporting the early use of intravenous thrombolysis in eligible acute stroke patients.

Bottom line: Expedient treatment with intravenous thrombolysis should occur in acute stroke patients.

Citation: Strbian D, Ringleb P, Michel P, et al. Ultra-early intravenous stroke thrombolysis: do all patients benefit similarly? Stroke. 2013;44(10):2913-2916.

Clinical Shorts


A large cohort study showed an increased rate of both hypo- and hyperglycemia in diabetic patients treated with fluoroquinolones vs. macrolides; of the fluoroquinolones used, moxifloxacin was the worst offender.

Citation: Chou HW, Wang JL, Chang CH, Lee JJ, Shau WY, Lai MS. Risk of severe dysglycemia among diabetic patients receiving levofloxacin, ciprofloxacin, or moxifloxacin in Taiwan. Clin Infect Dis. 2013;57(7):971–980.


Meta-analysis shows IV iron increases hemoglobin levels and reduces the need for red blood cell transfusion in patients with iron deficiency anemia. However, IV iron also increased risk of infection.

Citation: Litton E, Xiao J, Ho KM. Safety and efficacy of intravenous iron therapy in reducing requirement for allogeneic blood transfusion: systematic review and meta-analysis of randomized clinical trials. BMJ. 2013;347:f4822.


In patient surveys, placebos could be appropriate if no harm occurred and if the physicians gave honest opinions about placebos. Two thirds of patients would consider a placebo in some instances.

Citation: Hull SC, Colloca L, Avins A, et al. Patients’ attitudes about the use of placebo treatments: telephone survey. BMJ. 2013;347:f3757.


A metasynthesis of six reviews concludes that oral factor Xa inhibitors and direct thrombin inhibitors are effective after TKA-THA; compared to LMWH, the factor Xa inhibitors’ marginal clinical benefits are offset by their increased risk for bleeding.

Citation: Adam SS, McDuffie JR, Lachiewicz PF, Ortel TL, Williams JW. Comparative effectiveness of new oral anticoagulants and standard thromboprophylaxis in patients having total hip or knee replacement: a systematic review. Ann Intern Med. 2013;159(4):275-284.

Next Article:

   Comments ()