Show Summary Details
Page of

Preventing Hospital-Acquired Complications 

Preventing Hospital-Acquired Complications
Preventing Hospital-Acquired Complications

Jennifer Meddings

Page of

PRINTED FROM OXFORD MEDICINE ONLINE ( © Oxford University Press, 2022. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a title in Oxford Medicine Online for personal use (for details see Privacy Policy and Legal Notice).

date: 23 May 2022

  1. A. Introduction

    1. a. As many as 10% of hospitalized patients experience harm unrelated to an underlying illness during their inpatient stay, and approximately 2% of hospitalized patients experience either major permanent injury or death as a result of medical care.

    2. b. Most US hospitals are now mandated to measure and report rates of several types of hospital-acquired infections and noninfectious complications. Rates of such events now affect hospital payments and are publicly reported on a website called Hospital Compare (

    3. c. The term adverse event refers to an injury that results from medical care, not an underlying illness.

      1. i. An adverse event does not imply “error” or poor-quality care.

      2. ii. However, just as physicians target blood pressure control to reduce myocardial infarction and stroke, they can target errors to prevent adverse events.

  2. B. Definition and Prevention of Hospital-Acquired Complications

    1. a. Adverse drug event (ADE)

      1. i. Definition

        1. 1. An ADE refers to any adverse clinical effect resulting from a drug. The term includes effects ranging from minor drug rashes to major reactions such as bone marrow suppression or anaphylaxis. It also includes expected adverse reactions (i.e., side effects) that may occur when a drug is used in standard doses to achieve accepted therapeutic goals. Harm resulting from medication errors (e.g., dosing errors or ordering a medication for the wrong patient) are also included in such events.

        2. 2. Prevention of ADEs—both side effects and those due to errors—is important to ensure safety.

        3. 3. An ADE can reflect factors arising at any stage of the medication use process, although drug ordering accounts for approximately 50% of ADEs.

      2. ii. Incidence. At least 5% of hospitalized patients experience an ADE.

      3. iii. Prevention. Many ADEs can be anticipated and prevented

        1. 1. Medication information. Become familiar with a trusted resource for medication prescribing. Many such resources (e.g., Micromedex) are now readily accessible electronically. Use this resource to verify dose, route, and duration of administration as well as contraindications and possible adverse reactions. If such a resource is not available, seek the advice of a pharmacist before prescribing treatment.

          • a. For older adult patients or patients receiving multiple medications, review orders for potential drug interactions.

          • b. Pay particular attention to medications used to treat human immunodeficiency virus (HIV), immunosuppressant drugs, insulin, and anticoagulants such as heparin. Chemotherapy orders should only be written by physicians with special training and approval.

          • c. Be cautious with all medications in patients with renal or hepatic dysfunction. Reassess medication safety and dose appropriateness on a daily basis during hospitalization because renal and hepatic function can quickly change during hospitalization for acute illnesses.

          • d. During bedside rounds, review the medications being administered by intravenous pumps. Is the infusion rate and drug still appropriate for your patient?

          • e. Monitor for exacerbations of chronic medical illness that might occur because of failure to resume home medications (e.g., diuretics in patients with heart failure).

        2. 2. Ordering. Most hospitals require medication entry on an electronic medical record, a change that has reduced errors related to illegible handwriting and automated medication dosing safety checks. However, new problems and prescribing errors have emerged. For example, it is easy to enter the wrong prescription for the wrong patient on an electronic medical record when multitasking. Also, a high number of warnings from the electronic medical record can lead to “alert fatigue,” or ignoring pop-up boxes and other such alerts. Because errors during order entry may also occur, it is important to avoid abbreviations and pay close attention to decimals (Table 4.1).

          • a. Avoid abbreviations for drug names to reduce the risk for misinterpretation (e.g., HCTZ [hydrochlorothiazide] versus HCT [hydrocortisone]; MSO4 [morphine] versus MgSO4 [magnesium sulfate], CPZ [Compazine or chlorpromazine]).

          • b. Always write complete and specific orders. Do not drop units, omit frequencies, or leave route of administration unspecified (see Table 4.1).

          • c. Never write blanket orders such as “continue all pre-op meds.” Rather, review each medication to determine whether it should be discontinued, held, or resumed after surgery.

          • d. Always check drug allergies; do not assume someone else has done so. Verify the allergy list with the patient. Do not assume the list in the electronic medical record is complete or correct.

          • e. Always check the patient’s name at the top of the record (computer or chart based) before writing an order. It is very easy to become distracted and fail to recognize that the wrong patient’s record is open.

          • f. Do not assume the medication list in the electronic medical record is correct. For all new admissions, review the list of prescribed and over-the-counter medications (including vitamin/herbal supplements) with your patient or caregivers. Ask not only what was prescribed but also if and how the patient is taking the medication. For example, a patient who is prescribed multiple blood pressure medications but is not taking them may develop serious hypotension if these medications are administered together in the hospital.

            Table 4.1 Prescription Order Writing Example Do’s and Don’ts


            Do Not


            0.5 mg

            .5 mg

            The leading zero avoids a 10-fold overdose caused by an overlooked decimal point.

            5 mg

            5.0 mg

            Dropping the trailing zero avoids a 10-fold overdose caused by an over-looked decimal point.

            NPH Insulin 10 units

            NPH Insulin 10 U

            “U” can be misread as a tailing zero.

            Synthroid 100 mcg

            Synthroid 100 μ‎g

            “μ‎g” is more likely to be misread as “mg” for milligram

            Coumadin 5 mg PO every other day

            Coumadin 5 mg PO qod

            “qod” can easily be misread as “qd.” For handwritten orders, many sources recommend dropping all abbreviations for frequency (e.g., qd, bid) in favor of such terms as “once daily” and “twice daily”

            Hot Key

            Periodically check medication administration records to ensure that patients are receiving medications in the route and dose ordered. This is especially important when a patient is not responding to treatment or has an unexpected change in course. Be aware that some commonly prescribed medications, such as prophylaxis for venous thromboembolism, are often not administered as prescribed because of issues such as patient refusal.

        3. 3. Dispensing errors (e.g., mix-ups in the pharmacy that result in the wrong medication being sent to the floor or an incorrect concentration of an intravenous [IV] medication) are not uncommon. You should recognize that such mistakes can occur.

          • a. For example, if a patient’s blood sugar falls precipitously after their usual dose of insulin, check to make sure that a more concentrated form of insulin was not accidentally used.

          • b. Medications stored and dispensed in automated machines can end up in the wrong drawer, leading to dangerous drug substitutions.

          • c. Multi-use vials may be mistaken for one another (e.g., heparin and insulin look alike, and several deaths have been reported due to “line flushes” with insulin instead of heparin).

            Hot Key

            When evaluating an unexpected clinical event, include medication errors in your differential diagnosis.

        4. 4. Administration errors (such as programming errors for the IV pump administering patient-controlled analgesia) are often out of your control. But you should recognize that they occur.

          • a. Physician administration. Sometimes physicians will be called on to administer a drug because nurses are not permitted to administer it. Examples of such medications include IV adenosine, “IV pushes” of certain narcotics (e.g., fentanyl), and IV contrast for radiographic studies.

            Hot Key

            Do not view calls to administer medication (e.g., IV metoprolol on a general ward) as an annoyance. Make sure you know how to administer the medication, including speed of administration, need for central line versus peripheral line, compatibility with other medications, and responses to possible complications.

        5. 5. Anticipate preventable complications

          • a. Interactions with warfarin

            • i. Increased international normalization ratio (INR). Many medications interact with warfarin to increase the INR and bleeding risk. For instance, amiodarone and trimethoprim-sulfamethoxazole can both produce a substantial, sudden rise in INR and lead to serious bleeding. In such cases, increase the frequency and vigilance of anticoagulation monitoring.

            • ii. Medications that interact with warfarin can be remembered by the eight A’s: (1) most antibiotics; (2) antifungals (fluconazole, miconazole); (3) serotonergic antidepressants; (4) amiodarone; (5) acetaminophen; (6) antiplatelet agents; (7) alternative remedies (e.g., dong quai, fenugreek, chamomile, St. John’s wort, Gingko biloba). St. John’s wort (and rifampin) decrease the activity of warfarin. Such interactions represent preventable ADEs.

          • b. Steroids

            • i. Hyperglycemia commonly occurs in hospitalized patients with diabetes who are given steroids to treat illnesses such as asthma or chronic obstructive pulmonary disease (COPD). Hyperglycemia is usually predictable and can be preemptively managed by closer glucose monitoring and adjusting medications for diabetes in anticipation or hyperglycemia.

            • ii. Psychosis may occur in patients who receive high doses of steroids. Reassess the necessity and dose of steroid each day to minimize unnecessary use.

            • iii. Vertebral compression fractures in patients with COPD who take chronic steroids are preventable ADEs. Concomitant prescription of a bisphosphonate can significantly reduce the risk for fracture.

          • c. Common side effects of “as needed” medications in hospitalized patients. Medications used as needed (or prn) for sleep, itching, or constipation can have important side effects, particularly among older adults. For example, diphenhydramine can lead to urinary retention and delirium in older adults. Medications used to prevent or treat constipation can cause diarrhea and lead to unnecessary testing or isolation for Clostridium difficile infection.

    2. b. Venous thromboembolism (VTE)

      1. i. Definition. Nosocomial VTE includes lower extremity deep venous thrombosis (DVT), pulmonary embolism, and upper extremity DVT.

      2. ii. Incidence. VTE may be as common among medical as post-surgical patients. One recent study found that 15% of acutely ill medical patients experienced DVT. In particular, central venous catheters (CVCs) such as peripherally inserted central catheters (PICCs) can increase the risk for upper and lower extremity DVT. A systematic review found that PICCs are associated with 2.55 times greater risk for thrombosis than conventional “triple-lumen” catheters.

      3. iii. Prevention

        1. 1. Detailed guidelines outlining risk estimation and appropriate use of anticoagulants for prevention and treatment of VTE are available from the American College of Chest Physicians.

        2. 2. Risk estimation is a cornerstone of VTE prophylaxis. Commonly used scores to assess risk include the Caprini Risk Score (developed for surgical patients) and the Padua Risk Score (developed for both medical and surgical patients).

        3. 3. Common risk factors that increase the risk for VTE include presence of cancer, immobility, surgery during admission, and presence of a CVC such as a PICC.

    3. c. Falls

      1. i. Incidence. Inpatient falls are common. Hip fractures are one of the most feared complication of falls. Up to 20% of patients sustaining a hip fracture become nonambulatory, losing their ability to carry out activities of daily living.

      2. ii. Risk factors. Factors associated with hospitals such as unfamiliar rooms, improper bed height, and sedative medications may exacerbate risk. Acute illness itself and deconditioning from being bedridden also increase the risk for falls. Presence of handrails and no-slip bathing surfaces may mitigate the risk for falls or fall-related injury.

      3. iii. Prevention. Various strategies exist to assess patients at high risk for falls and prevent such events. Appropriate strategies for a given patient are individualized and often best discussed with nursing staff.

        1. 1. Avoid using sedating or disorienting medications.

        2. 2. Limit use of restraints and handrails—these may paradoxically increase risks for falls or the severity of fall-related injuries.

        3. 3. Urinary catheters do not reduce (and may increase) the risk for falls in patients with incontinence. Do not place urinary catheters as a strategy to reduce falls.

    4. d. Contrast-induced nephropathy (CIN)

      1. i. Definition

        1. 1. Specific definitions vary, but clinically significant CIN is commonly defined by a 25% or 0.5-mg/dL or more increase in creatinine above baseline within 48 hours of intravenous contrast administration.

        2. 2. The hallmark of CIN is that it resolves within 7–10 days and rarely requires specific therapy.

      2. ii. Incidence. Established risk factors for CIN include diabetes, chronic renal insufficiency, decreased effective circulating volume, and multiple myeloma.

        1. 1. For patients with no risk factors, the incidence of CIN approaches zero.

        2. 2. For patients with one or more of these risk factors, the incidence ranges from 5%–40%, depending on the number and severity of risk factors. CIN occurs in more than 50% of patients with a creatinine >4 mg/dL.

      3. iii. Treatment. Dialysis may be required but is often temporary. Supportive care and gentle hydration are cornerstones of treatment.

      4. iv. Prevention. CIN can be prevented by the following strategies:

        1. 1. Avoid drugs that decrease renal perfusion (e.g., nonsteroidal anti-inflammatory drugs [NSAIDs]).

        2. 2. Discontinue metformin before any dye-based procedure. In patients undergoing elective cardiac catheterization, metformin is discontinued at least 12 hours before the procedure.

        3. 3. Correct hypovolemia, especially in patients with diabetes or preexisting renal problems. Typical protocols for ensuring adequate volume status consist of normal saline administered at 75 mL/hr for 12 hours before and after the procedure.

        4. 4. For high-risk patients (e.g., diabetic patients with moderate renal insufficiency), use low-osmolar contrast agents and limit the volume of contrast (<75–125 mL). Repeated contrast exposures within 48–72 hours should be avoided.

        5. 5. For any patient with chronic renal insufficiency, consider N-acetylcysteine treatment; although the magnitude of its effectiveness to prevent CIN varies by studies, it is inexpensive and well-tolerated. The drug is administered as powder-containing capsules or diluted in soda or juice. A common dosing schedule for this medication to prevent CIN is to give 600 mg orally twice daily on the day before and after the procedure. If only one dose is given before the procedure, give three doses afterward.

        6. 6. Recent systematic reviews have questioned the benefit of IV sodium bicarbonate to prevent CIN.

    5. e. Stress-related upper gastrointestinal bleeding (UGIB)

      1. i. Incidence

        1. 1. Clinically significant UGIB occurs in 0.1% of low-risk intensive care unit (ICU) patients and in approximately 3% of ventilated patients.

        2. 2. General improvements in ICU care and focus on clinically significant events have reduced the frequency of stress-induced gastric ulceration among ICU patients.

      2. ii. Prevention. Although a benefit of pharmacologic agents that suppress gastric acid (H2-receptor antagonists, proton pump inhibitors) has been shown in some studies, the overall clinical benefit of stress-ulcer prophylaxis in unselected patients is unclear. Because these medications are also associated with increased risk for nosocomial pneumonia, C. difficile infection, and drug-specific ADEs such as thrombocytopenia, routine use is not recommended.

        Hot Key

        Do not routinely prescribe antacid medications to ICU patients. Focus on patients with risk factors for UGIB: respiratory failure, coagulopathy, renal failure, and significant burns.

    6. f. Catheter-associated urinary tract infection (CAUTI)

      1. i. Incidence. Catheterized patients develop bacteriuria at a rate of 3%–10% per day. Among patients with bacteriuria, about 10% will develop UTI symptoms such as dysuria or flank pain, and approximately 3% develop bacteremia.

        Hot Key

        Make sure you know which of your patients have a urinary catheter and why they still need one. Order urinary catheters only when the patient meets appropriate clinical use criteria, use alternatives to indwelling urinary catheters when possible, and assess continued appropriateness on a daily basis.

      2. ii. Prevention

        1. 1. Not only may urinary catheters lead to infection, they may also be a source of significant discomfort and embarrassment to patients. Remove when no longer clinically necessary.

        2. 2. Condom catheters may lower the risk for UTIs in certain patients and may be more comfortable for males.

        3. 3. Antimicrobial catheters (e.g., silver alloy–coated catheters) are no longer supported as effective strategies for preventing CAUTI.

        4. 4. Do not send urine cultures for patients without clinical signs of UTI. Bacteria in the urine is not equivalent to a UTI. Bacteriuria is common and often unavoidable among patients with indwelling catheters. Risk factors for asymptomatic bacteriuria include advanced age, postmenopausal status, and diabetes. Asymptomatic bacteriuria is recommended in specific cases (pregnancy, patients undergoing urologic procedures with anticipated mucosal injury). Changes in urine odor and urine color are not symptoms of infection and do not warrant testing.

    7. g. Pressure injury to skin (known as pressure ulcers, bedsores, or decubitus ulcers)

      1. i. Incidence

        1. 1. Approximately 5% of hospitalized patients have pressure-related skin injuries. Most (60%) of these develop in the hospital.

        2. 2. The prevalence of pressure injury is much higher in immobilized patients; about 30% of older adult patients with a fractured hip develop pressure ulcers in the preoperative period.

      2. ii. Prevention

        1. 1. Inspect all hospitalized patients for early signs of pressure ulcers such as erythema over bony prominences that contact the hospital bed (e.g., occiput, scapula, elbow, sacrum, greater trochanter, ischial tuberosity, malleoli, heel), skin that rubs in immobile patients (such as between knees), and skin that is in contact with common medical devices such as oxygen tubing or masks (e.g., behind ears, under nose, lips). Additionally, urinary catheters behind legs or mechanical compression devices to prevent VTE can also cause skin injury.

        2. 2. Identify patients at risk for development of pressure ulcers. Immobility (e.g., due to neurologic disorder or fracture), malnutrition, fecal or urinary incontinence, and altered level of consciousness) are risk factors for skin breakdown. At-risk patients should have pillows placed beneath the calves, foam foot protection, or a soft foam or air waffle mattress to minimize injury.

        3. 3. Patients unable to move independently in bed should be “turned” or repositioned every 1–2 hours.

    8. h. Central line-associated bloodstream infection (CLABSI)

      1. i. Incidence. Among patients with a standard CVC in place for 1 week or more, approximately 25% develop catheter colonization and 5% become bacteremic. CLABSI occurs at a rate of about 5 per 1000 catheter-days in medical and surgical ICUs. Catheter-associated VTE is also a common complication of CVC use, with higher rates for PICCs than CVCs.

      2. ii. Prevention

        1. 1. Order and place CVCs (including PICCs) only when the patient has an appropriate clinical indication. Use alternative strategies for medication delivery to avoid complications related to PICC or CVC use.

        2. 2. Use maximum sterile barriers when inserting CVCs. This includes sterile gloves, a long-sleeved gown, a full-sized drape, and a nonsterile mask cap.

        3. 3. Use alcohol-containing chlorhexidine gluconate as opposed to povidone-iodine for skin antisepsis over the insertion site.

        4. 4. The subclavian site has the lowest risk for infection compared with internal jugular or femoral insertion.

        5. 5. Remove the CVC or PICC as soon as clinically possible. Do not continue with the CVC simply for convenience, especially if the patient is going to be transferred to a regular floor, where knowledge of CVC care is variable.

        6. 6. Use a clinical risk calculator to determine the risk for CLABSI associated with a PICC.

          Hot Key

          Always place the patient in the supine position when removing a CVC from the subclavian or internal jugular sites to avoid air embolism.

    9. i. C. difficile infection is a common, costly, and morbid infection. It can be either a community-acquired or hospital-acquired infection.

      1. i. Incidence. Asymptomatic C. difficile colonization occurs in both healthy and hospitalized adults. Colonized adults have a sixfold greater rate of clinical infection when hospitalized.

      2. ii. Diagnosis. Clinical manifestations of C. difficile infection include diarrhea with colitis (typically three or more loose stools in a 24-hour period), leukocytosis, ileus, and symptoms such as nausea, vomiting, abdominal pain, cramping, and fever. Laboratory tests available for diagnosis include stool tests for the C. difficile antigen and toxin. The preferred testing technique varies by hospital and should be discussed with your microbiology laboratory.

      3. iii. Prevention. The most important strategy for preventing C. difficile is reducing antibiotic use. Avoiding antibiotics in the absence of clear benefits (e.g., viral respiratory infections) and using antibiotics only for as long as clinically necessary are recommended. The evidence for probiotics in preventing C. difficile is mixed. Routine use of probiotics to prevent or treat C. difficile is not recommended.

    10. j. Hospital-acquired and ventilator-associated pneumonia: All patients, including those who are mechanically ventilated, are at risk for developing pneumonia during their hospitalization.

      1. i. Incidence. Hospital-acquired pneumonia and ventilator-associated pneumonia are believed to cause about 20% of hospital-acquired infections, with a mortality rate of 10%–15%.

      2. ii. Ventilator-associated events (VAEs). Patients who require mechanical ventilation are at increased risk for infectious (e.g., ventilator-associated pneumonia [VAP]) and noninfectious (e.g., lung injury from barotrauma, including pneumothorax) complications.

        1. 1. Diagnosis. One type of VAE, VAP, is difficult to diagnose. In clinical practice the diagnosis is often made by radiographic findings or positive endotracheal, pleural, or blood cultures in the setting of fever, leukocytosis, or purulent tracheal secretions. The effect of VAP on clinical outcomes varies depending on the organisms involved and comorbid conditions.

        2. 2. Prevention. Several strategies to prevent VAP exist. Avoiding intubation by using noninvasive positive pressure ventilation is a key strategy. Additionally, use of bundles that include the following have been associated with reduced rates of VAP:

          • a. Oral care with chlorhexidine

          • b. Daily sedation interruption

          • c. Daily assessment of readiness to extubate

          • d. Minimal interruptions of the ventilator circuit

          • e. Intermittent or continuous aspiration of subglottic secretions and selective digestive tract decontamination by nonabsorbable antibiotics

        3. 3. Early physical mobility of intubated patients is also associated with lower rates of VAP. While elevation of head of bed to 30°–45° is commonly recommended, limited data support this recommendation.

  3. C. General Errors. Although the complications described in B should be considered in the care of hospitalized patients, more general opportunities for serious errors and complications should also be kept in mind.

    1. a. Pay attention to internal warning signals.

      1. i. Haste leads to mistakes. If you begin to feel overwhelmed, rushed, or pressed for time (e.g., you cannot possibly handle all of your work), step back and take a few minutes to gather your thoughts.

      2. ii. Ask for help. The best learners know when they are in over their head and are not afraid to ask for help.

    2. b. Communicate often with nurses, pharmacists, and physicians from other services involved in your patients’ care. Both minor and major errors (e.g., taking the wrong patient for a procedure) occur because of poor communication among physicians or between physicians and nurses.

      1. i. Communicate important changes in patient care plans to all clinical personnel, including nurses, physical therapists, pharmacists, and clinical consultants. Do not forget to include the patient’s primary care physician at the time of discharge.

      2. ii. Conversely, if you are surprised to find a new order or care plan for your patient, do not assume that a clinician has simply devised a new plan and not told you. Check out the new plan yourself.

    3. c. Communicate clearly with patients in a manner they understand. Nearly 25% of adults in the United States are functionally illiterate, and even more have inadequate health literacy. More than 50% of patients seen in public hospitals do not understand common medical terms such as “malignant,” “terminal,” or even “orally.” Basic medication instructions such as “take every 6 hours” or “take on an empty stomach” are poorly understood by at least 25% of patients, and more than 50% cannot reliably interpret a blood sugar reading.

      1. i. Slow down and use “living room” language instead of medical terminology; show or draw pictures whenever possible.

      2. ii. Ask, “Can you tell me what we have just gone through?” or other “teach back” approaches to confirm understanding.

      3. iii. Limit information conveyed during a single interaction. Do not overwhelm patients with information, especially in the emergency department, ICU, or a first-time meeting in the middle of the night.

    4. d. Error disclosure. When an error occurs, communication of the error, issuance of an apology, and taking steps to correct the problem so that it does not recur are important. Trainees who recognize errors should promptly communicate the concern to their supervising physician who will lead the team for disclosing the error and apologizing to the patient. Disclosure can strengthen patient trust in the patient-physician relationship.

Suggested Further Readings

Chopra V, Flanders SA, Saint S, et al. The Michigan Appropriateness Guide for Intravenous Catheters (MAGIC): results from a multispecialty panel using the RAND/UCLA Appropriateness Method. Ann Intern Med 2015;163:S1–40.Find this resource:

Gallagher TH, Studdert D, Levinson W. Disclosing harmful medical errors to patients. N Engl J Med 2007;356:2713–9.Find this resource:

Guyatt GH, Akl EA, Crowther M, Gutterman DD, Schunemann HJ. Executive summary: Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines. Chest 2012;141:7s–47s.Find this resource:

Kachalia A, Kaufman SR, Boothman R, et al. Liability claims and costs before and after implementation of a medical error disclosure program. Ann Intern Med 2010;153:213–21.Find this resource:

Kalil AC, Metersky ML, Klompas M, et al. Management of adults with hospital-acquired and ventilator-associated pneumonia: 2016 clinical practice guidelines by the Infectious Diseases Society of America and the American Thoracic Society. Clin Infect Dis 2016;63:e61–e111.Find this resource:

Meddings J, Saint S, Fowler KE, et al. The Ann Arbor criteria for appropriate urinary catheter use in hospitalized medical patients: results obtained by using the RAND/UCLA appropriateness method. Ann Intern Med 2015;162:S1–34.Find this resource:

Pronovost P, Needham D, Berenholtz S, et al. An intervention to decrease catheter-related bloodstream infections in the ICU. N Engl J Med 2006;355:2725–32.Find this resource:

Saint S, Greene MT, Krein SL, et al. A program to prevent catheter-associated urinary tract infection in acute care. N Engl J Med 2016;374:2111–9.Find this resource:

Subramaniam RM, Suarez-Cuervo C, Wilson RF, et al. Effectiveness of prevention strategies for contrast-induced nephropathy: a systematic review and meta-analysis. Ann Intern Med 2016;164:406–16.Find this resource: