Evidence: Torsemide has better oral bioavailability and a longer half-life than furosemide. Metanalysis of 9 RCT and 11 observational studies compared 4,550 patients with heart failure treated with torsemide vs 14,730 patients treated with furosemide for a mean follow up duration of 15 months. Findings showed that torsemide was associated with significant improvement in functional status in terms of NYHA class (72.5% vs 58%, OR 2.32, NNT = 5). Cardiac mortality was lower in the torsemide group although all-cause mortality was not different. Torsemide was also associated with a numerically lower risk of hospitalization for heart failure (not statistically significant) with no difference in side effects. One of the limitations was the high degree of heterogeneity among the studies. The positive finding on cardiac mortality should be interpreted with caution as none of the studies were powered to assess cardiac mortality and historically no mortality benefit of diuretic therapy has been demonstrated.

Pearl: Consider switching diuretic therapy from furosemide to torsemide in patients with heart failure, especially for patients with increased NYHA class or repeated hospitalizations due to heart failure exacerbation.

Evidence: A multimodal approach to pain control in the postoperative setting, which includes the use of acetaminophen and NSAIDs, has been suggested to improve pain control and reduce opioid use. The aim of the randomized, double-blinded PANSAID (Paracetamol and NSAID in combination) trial was to investigate the analgesic (morphine sparing) and harmful effects of 4 multimodal analgesic regimens with acetaminophen, ibuprofen, or both in combination after total hip arthroplasty (THA). Patients received either combined acetaminophen 1000 mg and ibuprofen 400 mg (n=136), acetaminophen + placebo (n=142), ibuprofen + placebo (n=140) or combined half strength acetaminophen (500 mg) + ibuprofen (200 mg) (n=140) every 6 hours for 24 hours post-operatively. The 2 co–primary outcomes were 24-hour morphine consumption using patient-controlled analgesia (PCA) and proportion of patients with 1 or more serious adverse events (SAEs) within 90 days after surgery without prolonging length of hospitalization. The minimal clinically important difference (MCID) was predetermined to be morphine 10 mg. When compared with acetaminophen alone, the combined acetaminophen and ibuprofen group used significantly less morphine (20 mg vs 36 mg; p < 0.001). There was no clinically significant difference in morphine use between the other compared groups. There were no significant differences in rates of serious adverse events in patients randomized to ibuprofen compared with acetaminophen 90 days postoperatively. A limitation of this trial was the intervention period was only 24 hours postoperatively making generalizability to our patient population limited as post-operative patients require longer duration of therapy.

Pearl: Consider using acetaminophen 1000 mg and ibuprofen 400 mg every 6 hours for the first 24 hours postoperatively in patients undergoing total hip arthroplasty. This regimen also resulted in lower pain scores at rest at 24 hours and a lower risk of nausea at 24 hours postoperatively (secondary and exploratory outcomes). Caution that the 24-hour intervention period is a major limitation of this study, as the majority of our patients require a longer duration of analgesia post-operatively.

Evidence: Vascular access becomes a lifeline for patients with advanced chronic kidney disease (CKD), who may eventually need renal replacement therapy. In 2003, CMS and the ESRD networks jointly formed and implemented a National Vascular Access Improvement Initiative called the Fistula First Initiative (FFI). Arteriovenous fistula (AVF) is the preferred choice for long-term hemodialysis with lower rates of complications and mortality when compared to tunneled dialysis catheters. The success of AVF placement often depends on vein preservation because frequent manipulations of the native veins such as placing peripheral IV’s or centrally placed catheters (ie peripherally inserted central catheter (PICC)) have been associated with an increased risk for venous thrombosis and central vein stenosis. The recommendation is that PICC placement should be avoided in patients with stage 3b (eGFR < 45 mL/min/1.73 m2) or greater CKD. One explanation for the widespread use of PICCs in patients with CKD is a knowledge gap regarding the appropriate use of PICCs. A large, prospective cohort, multicenter study focused on the frequency and characteristics of PICC use in patients with CKD was conducted across 52 hospitals participating in Michigan from November 2013 to September 2016. The primary outcome was the percentage of patients with stage 3b or greater CKD receiving a PICC. Secondary outcomes were PICC-related complications and central line–associated infections. Out of 20,545 patients who had PICCs placed, 23.1% had stage 3b or greater CKD and 3.4% were receiving hemodialysis. The proportion of PICCs placed in patients with CKD increased significantly with age in both ICU and ward settings. 17.8% had PICC-related complication and the most common complication was catheter occlusion. Complications, including VTE, were more common among patients with CKD with multi-lumen PICCs than among those with single-lumen PICCs.

Pearl: Avoid PICCs in patients with stage 3b or greater CKD for vein preservation, in preparation for possible future dialysis access placement. Multi-lumen PICCs have more complications than single lumen PICCS. The American Society of Nephrology recommends consultation with a nephrologist before placement of a PICC in patients with CKD stage 3 to 5.

Evidence: Identification and coding of diabetic complications directly affects risk adjustment under the hierarchical condition categories (HCCs) that influence Medicare quality metrics. Diabetic complications need to be specifically identified and coded because the HCCs for these codes have a much higher value than the HCC for uncomplicated diabetes. ICD-10-CM only recognizes four acute diabetic complications: hypoglycemia, hyperglycemia, hyperglycemic hyperosmolar state (HHS), and diabetic ketoacidosis (DKA). The Pearls team will cover the five chronic complications in issue 16. Hypoglycemia is defined as a blood glucose level below 70 mg/dL with codes for the presence or absence of concomitant coma (even if transient). Hyperglycemia is defined as a blood glucose level above 140 mg/dL. Whether the cause is hypoglycemia or hyperglycemia, the Glasgow Coma Scale score should be recorded for all cases with coma or with any changes in level of consciousness. Under ICD-9-CM, the distinction between controlled and uncontrolled diabetes was important. With ICD-10-CM, these terms are no longer used to describe diabetes. The term “uncontrolled” must be clarified by clinicians as either hyperglycemia or hypoglycemia. Codes for nondiabetic hyperglycemia and hypoglycemia also exist but are never used with diabetes. HHS and DKA are commonly overly diagnosed and improperly coded. HHS is defined by the presence of both a blood glucose level above 250 mg/dL and serum osmolality greater than 320 mmol/L. The diagnosis of DKA requires all of the following: 1) a blood glucose level above 250 mg/dL 2) acidosis with a pH below 7.30; 3) a bicarbonate level below 18 mEq/L; and 4) markedly elevated serum ketones. Therefore, DKA cannot be properly diagnosed without an arterial blood gas and serum ketone measurement.

Pearl: There are only four types of acute diabetic complications: hypoglycemia, hyperglycemia, DKA, and HHS. Document these complications as they directly contribute to risk adjustment. Many patients will experience hyperglycemia during hospitalization, so document and bill for hyperglycemia on days when it occurs. Avoid documentation and billing of hyperglycemia on days when it is absent. Terms such as controlled or uncontrolled diabetes should no longer be used. Order a serum osmolality before documenting HHS. Order an arterial blood gas and serum ketones before documenting DKA.

Evidence: As social media continues to develop, medicine is trying to keep up. The average medical professional has limited bandwidth to seek out and digest ongoing research and thus relies on other medical experts for guidance. There are numerous leaders in medicine who are flipping the script on knowledge delivery and have created excellent curriculums via platforms like podcasting and twitter. A popular podcast is “The Curbsiders,” whose mission is to “curbside the experts to deconstruct various topics in the world of medicine to provide listeners with clinical pearls, practice changing knowledge, and bad puns.” This episode features a regular to the podcast, Dr. Joel Topf (nephrologist and twitter guru @kidney_boy), who educates us on AKI. The PIMG Pearls team will continue to reference prominent social media figures who are challenging the way we continue to learn.

Pearl: KDIGO criteria for AKI is primarily for research and is not a practical definition for our clinical practice (consider the aggressively resuscitated trauma patient who is oliguric but their creatine may be stable). Documenting oliguria and/or anuria is critical to risk stratify AKI. If a patient is anuric and receiving fluids, watch out because a drop in creatinine is likely dilutional. Rhabdomyolysis and contrast induced nephropathy are two intrinsic renal injuries that can have a pre-renal FENa (<1%). In rhabdomyolysis, it is usually safe to discharge reliable patients on oral fluids once the creatinine and creatinine kinase have peaked with planned repeat outpatient labs. No guidelines to wait for a specific halving threshold to safely discharge. Approximately 90% of patients will improve with IV fluids and a foley catheter, and those that don’t improve likely have ATN which will get better with time.

Evidence: Ammonia is produced by colonic bacteria and enters the blood via the portal vein. It is normally metabolized by the liver (as well as muscle and brain), however, in liver dysfunction, ammonia metabolism is impaired. Ammonia is known to be toxic to the brain and is thought to be involved, at least some of the time, in causing the clinical syndrome of hepatic encephalopathy. Ammonia levels are commonly elevated in patients with chronic liver disease suspected to have hepatic encephalopathy. However, there is a mountain of evidence to suggest that ammonia levels have poor predictive value for hepatic encephalopathy in patients with chronic liver disease. In this population, elevated serum ammonia correlates with hepatic encephalopathy, but it is neither sensitive enough to rule out hepatic encephalopathy, nor specific enough to rule in hepatic encephalopathy. Checking ammonia levels were identified by the Choosing Wisely: TWDFNR campaign as a low value intervention as it does not add diagnostic, staging, or prognostic value in chronic liver disease patients with hepatic encephalopathy.

Pearl: Do not check serum ammonia levels in patients with CLD to diagnose HE, to assess the severity of HE, or to determine whether HE is resolving. Use your clinical evaluation to determine the severity and course of HE using staging systems like the West Haven Criteria. Treatment should be tailored according to clinical findings, not ammonia levels.

Evidence: An analysis of 75,000 hospitalized patients aged 18-99 with cirrhosis and ascites who underwent paracentesis looking at timing of paracentesis (primary outcome) and rates of inpatient mortality, 30-day mortality, and SBP incidence. Only 58% of patients met the primary outcome of early paracentesis defined as paracentesis within 24 hours of admission. Those patients that underwent early paracentesis had significantly lower rates of all-cause inpatient mortality (7% vs. 10%) and 30-day readmissions than did patients who underwent late paracentesis. High-risk patients (hepatic encephalopathy, AKI or both) were significantly less likely to receive early paracentesis than usual-risk although more likely to have SBP.

Pearl: Early paracentesis within 24 hours of hospital admission is associated with reduced inpatient mortality, SBP related mortality and fewer 30-day readmissions. This mortality benefit may be related to early initiation of evidence-based therapy for SBP. On admission, consider asking the ED physician to do a diagnostic paracentesis as part of the initial work-up.

Evidence: Digoxin is FDA approved to treat HF and has been shown to reduce the risk of heart failure related hospitalizations and emergency care. Per ACC/AHA guidelines, digoxin can be used to decrease hospitalizations in patients with HFrEF, although it has never been shown to have a mortality benefit (see digitalis investigation group (DIG) trial). This JACC study used a propensity score-matched cohort of hospitalized patients with HFrEF (<45%) from the OPTIMIZE-HF, a national web-based registration of heart failure hospitalizations from hospitals across 48 states. Besides digoxin, patients were also on ACEI, BB, and mineralocorticoid receptor blockers. The outcomes of the study included HF readmission, all-cause readmission, all-cause mortality, and the combined endpoint of HF readmission or all-cause mortality and they were examined at 30 days, 6 months, 1 year and 4 years after discharge. Discontinuation of digoxin was associated with higher readmissions at 6 months, lasting up to 4 years but not at 30 days. Among patients with HFrEF who are on ACEI/ARB, beta-blocker and mineralocorticoid receptor blockers, discontinuing pre-admission digoxin was associated with higher risks of heart failure readmissions, but did not benefit all-cause mortality.

Pearl: Discontinuing digoxin in a patient who was previously tolerating the drug after hospitalization for HF exacerbation is associated with an increased risk of readmission. This study was not set up to demonstrate digoxin’s effectiveness in the digoxin-naïve HFrEF patient. There is a lack of evidence to support mortality benefit with digoxin.

Evidence: Acute severe ulcerative colitis (ASUC) is defined as > 6 bloody bowel movements and fever, tachycardia, anemia (Hgb < 10.5) or elevated inflammatory markers (CRP > 30). Patients require admission for initiation of IV corticosteroids or rescue therapy after failing outpatient management with oral corticosteroids. Patients with ASUC should be managed in consultation with a gastroenterologist. Triggers for ASUC include NSAIDs, recent antibiotic use, and infection including C difficile. In addition to a BMP, CBC and stool studies to rule out infection, laboratory evaluation should include a CRP for trending response to treatment. Quantiferon Gold, Hepatitis B serology, and thiopurine methyltransferase (TPMT) enzyme should be considered if patients fail IV corticosteroids and require infliximab or thiopurine use since one third of patients do not respond to initial IV therapy. CT imaging is not recommended due to recurrent radiation exposure; however, abdominal X-ray is recommended to assess for bowel dilatation. Flexible sigmoidoscopy can assess disease severity and rule out superimposed CMV colitis and is typically pursed if patients are not responding to IV corticosteroids after 3 days. All patients with ASUC should be prescribed VTE prophylaxis and it is not contraindicated in patients with rectal bleeding because of the increased risk of VTE in patients with inflammatory bowel disease. Anti-motility agents, opioids, NSAIDs and antibiotics (in the absence of infection) should be avoided. Methylprednisolone should be initiated at a dose of 20 mg IV every 8 hours and if there is an inadequate response (decrease to < 4 non-bloody BM per day) then rescue therapy with infliximab or cyclosporine or surgical consultation for colectomy should be discussed with a consulting gastroenterologist. Patients who respond to steroids can transition to oral prednisone 40 – 60 mg daily with planned taper and gastroenterology follow up when frequency of bowel movements has improved (goal 1-2 per day) and they are tolerating a regular diet.

Pearl: Acute severe ulcerative colitis should be managed in consultation with a gastroenterologist. Patients should be started on IV corticosteroids and assessed daily for response to treatment by monitoring bowel movement frequency and CRP. VTE prophylaxis should be ordered for all patients due to increased risk for VTE. Anti-motility agents, opioids, and routine use of antibiotics should be avoided. In steroid refractory disease, discuss initiation of immunosuppressive therapy or colectomy with the consulting gastroenterologist and surgeon if there is no response after initial 72 hours.

Evidence: On June 30th 2020, the European Delirium Association held a panel review of delirium in their COVID-19 patients. The speakers were from hospitals in England, Italy, and Ireland, and here are a few pearls that are clinically relevant:

Pearl: There is an increased rate of delirium in older patients with COVID-19 in the community, hospitals, and nursing homes when compared to their younger counterpart. Older patients with COVID-19 are more likely to die, but older patients with COVID-19 who present with delirium or develop delirium during their hospitalization are even more likely to die Oxygen use may be helpful in preventing delirium, even if the patient doesn't appear to be hypoxemic initially. The increased rates of delirium may be attributed to hypoxia and/or hypotension and/or thromboses, though direct neuro-invasion is certainly being entertained, as there are ACE2 receptors found in the brain. One thought on the neuro-invasion is possibly via the nasal epithelium. When MRIs are performed on these patients, microhemorrhages are often detected. When EEGs are performed, there is diffuse slowing. PET scans show reduced glucose metabolism. There have not been enough LPs performed to comment on whether the virus is detected in the CSF. Challenges in combating delirium in COVID-19 patients include inability to provide non-medical treatment such as family visitation or group therapy. Other challenges include limited access to testing such as MRI, EEG, and postmortem brain biopsies due to fear of exposing those performing the tests to COVID-19. Remember that hypoactive delirium is associated with a higher morbidity and mortality than hyperactive delirium, and hypoactive delirium is usually underrecognized.

Evidence: This update of the 2007 community-acquired pneumonia (CAP) guidelines involves a more evidence-based approach and includes 16 management questions addressing initial diagnosis through therapy completion and follow-up. I have summarized the recommendations that are the most relevant to our daily inpatient practice.

Pearl: Obtain Legionella and pneumococcal urinary antigen assays only in patients with severe CAP or epidemiologic risk factors. Do not routinely add anaerobic coverage for suspected aspiration pneumonia unless lung abscess or empyema is suspected. Do not use serum procalcitonin levels to determine initiation or duration of antibiotic therapy in patients with radiographically confirmed CAP unless antibiotic therapy is being extended beyond 5 to 7 days. Do not routinely treat with corticosteroids except in patients with concurrent refractory septic shock. Recommendations have been expanded for collection of sputum and blood cultures from patients with severe disease to also include any inpatient receiving empiric coverage for methicillin-resistant Staphylococcus aureus (MRSA) or Pseudomonas aeruginosa. The concept of healthcare-associated pneumonia (HCAP) has been eliminated. Providing empiric antibiotic coverage for MRSA or P. aeruginosa is recommended only if there are locally validated risk factors for either pathogen. If local data are not available, before starting such coverage, cultures of blood and lower respiratory secretions should be obtained to guide the need for continuing that treatment. For severe CAP, use of a β-lactam agent plus a macrolide is now favored over a β-lactam agent and a respiratory fluoroquinolone.

Evidence: Surgical antimicrobial prophylaxis is recommended for 24 hours postoperatively for most procedures. This recommendation is based on prior studies that demonstrated that longer courses did not further decrease risk of infection and may increase risk of antibiotic associated adverse events. In this VA retrospective cohort study of 79058 surgical patients, increased surgical prophylaxis duration (24 h - 48 h, 48 h - 72 h, > 72 h) was associated with increased risk of acute kidney injury (unadjusted NNH 9, 6, 4 respectively) and C diff infections (unadjusted NNH 2000, 90, 50). The risk increased with each additional day of antibiotic exposure. Surgical prophylaxis for > 24 hours was not associated with a reduction in surgical site infections.

Pearl: Longer duration (> 24 hours) of surgical antimicrobial prophylaxis was associated with duration dependent increased risk of antibiotic associated adverse events including acute kidney injury and C diff infections but not with reduction in surgical site infections. When consulting on surgical patients, consider discussion with the surgeon to limit the duration of antimicrobial prophylaxis to <24 hours.

Evidence: Hospitalists commonly document a troponin elevation NOT due to an MI as troponin leak or troponinemia which are non-specific descriptions. Occasionally these troponin elevations are documented improperly as type 2 myocardial infarctions. When we mislabel a diagnosis, it can impact billing and reimbursement, DRG denials, insurance premiums and quality metrics for both the hospital and the physicians. The Fourth Universal Definition of MI published in August 2018 further updated the definitions of MI. Type 1 and type 2 MIs are the most common types encountered by hospitalists unlike types 3-5 (listed below). As hospitalists, we are much more comfortable diagnosing and documenting a type 1 MI, so here we will expand on type 2 MIs and non-MI troponin elevations due to an underlying cause. When there is ONLY elevated troponin levels (or even a rise and fall in troponin) WITHOUT new symptoms or ECG/imaging evidence of myocardial ischemia, it is most appropriate to document a non-MI troponin elevation due to a nonischemic mechanism of myocardial injury. Causes may overlap with type 2 MI and include acute on chronic systolic or diastolic heart failure, pericarditis and myocarditis, stress induced cardiomyopathy (takotsubo), acute PE, sepsis without septic shock, renal failure (acute on chronic kidney disease/ESRD), and stroke.

Pearl: Mislabeling a diagnosis can impact billing and reimbursement, DRG denials, insurance premiums, and quality metrics for both the hospital and the physicians. As a spitalist – distinguishing between a type 2 MI and non-MI related troponin elevation is critical.

Evidence: Thrombocytopenia is common in hospitalized patients and often due to an underlying illness such as sepsis, acute hemorrhage, and liver disease, or can be the result of a medication effect. Heparin-induced thrombocytopenia (HIT) is an uncommon cause of thrombocytopenia with an incidence of 0.2 – 5%. Testing for HIT with HIT ELISA, which detects anti-PF4 antibody, should be reserved for patients with intermediate or high probability for HIT based on the 4T’s Scoring system (see below). If there is an intermediate or high probability (≥ 4), heparin should be held and a non-heparin anticoagulant should be initiated to prevent thrombotic complications while awaiting results of confirmatory testing. A low probability 4T’s score (≤ 3) has a negative predictive value of 99%, therefore, HIT ELISA testing is unnecessary and may lead to false positive results due to its high sensitivity. The resulting false positives in patients with a low pretest probability leads to unnecessary further diagnostic testing with the confirmatory serotonin release assay (SRA) and inappropriate treatment with non-heparin anticoagulation.

Pearl: In patients with low probability of HIT (4T’s score ≤ 3), consider alternative causes of thrombocytopenia and avoid ordering HIT ELISA testing.

Evidence: Inpatient diabetes care is getting more complicated due to the array of new oral drugs and insulin formulations, in addition to the limited availability of endocrinology consultation services. We will present an inpatient scenario whose clinical pearl is high yield to our daily practice. Consider an example involving long-acting insulin degludec injection, which has a half-life of 24 hours and can remain in the body as long as 42 hours. A patient with diabetes is admitted and takes 40 units of degludec every morning. But the hospital only has glargine available. So the hospitalist prescribes 40 units of glargine (which has a half-life of 12 hours) for the next morning at 8 a.m. Twelve hours after that first dose of glargine, the patient still has half of the glargine units, roughly 20, on board. Those units combined with the remaining 30 units or so of degludec result in a total of 50 units of circulating basal insulin by that evening. This concept is called insulin stacking which occurs if a patient is on a specific formulation of long acting insulin as an outpatient and is then initiated on long acting insulin as an inpatient without regard to the half-life of the patient’s home insulin regimen, which can lead to dangerous hypoglycemic events.

Pearl: Insulin stacking can cause serious hypoglycemic events. Consider the half-life of long acting insulins and if possible request the patient’s family member or friend to bring non-formulary insulins to the hospital.

Evidence: It has been apparent through employee monitoring that many health care workers, despite using personal protective equipment (PPE) on a daily basis, do not safely and effectively put on and take off PPE. Most health care workers do not know that the act of putting on and taking off PPE is known as donning and doffing.

Pearl: Donning and Doffing PPE Refresher Video (Start 5 min mark)

Evidence: For PCR testing, BAL and sputum culture are the most accurate, but there is concern for aerosolization of virus with bronchoscopy or sputum induction.Nasopharyngeal swabs are more sensitive than oropharyngeal swabs. The precise study characteristics of current PCR testing are unknown. However, a study published from China showed ~ 73% sensitivity. The positive predictive value is very high but there will be false negatives. Nasopharyngeal swabs are more sensitive than oropharyngeal swabs. The precise study characteristics of current PCR testing are unknown. However, a study published from China showed ~ 73% sensitivity. The positive predictive value is very high but there will be false negatives. CT scans are useful for identifying early disease with findings of ground glass opacities progressing to multifocal consolidation. CT imaging has a higher sensitivity than the PCR studies. 97% of SARS-CoV-2 RT-PCR positive patients had characteristic CTs. RT-PCR by throat swab was positive in 59% of the patients in this 1,014 patient series. For RT-PCR negative but positive chest CT scans, 48% were considered highly likely to have COVID-19. There were several patients who had initial positive CT and initial negative RT-PCR who later showed positive RT-PCR. CT scans were also useful for tracking the clinical course of the disease. 42% showed improvement in follow-up chest CT scans before RT-PCR became negative.

Pearl: Consider keeping patients with high clinical suspicion on droplet precautions even if they test negative. Consider CT imaging and repeating RT-PCR testing if not improving.

Evidence: Procalcitonin: SARS-CoV-2 is a procalcitonin negative viral infection and a high procalcitonin should raise the alarm for bacterial super-infection. CBC with differential for lymphocyte count: Bronchoscopy and autopsy testing of the lungs show high rates of lymphocytes in the lungs. The largest (1099 patients) series published in the NEJM from China showed 80% of patients with mild cases and 96% with severe cases had lymphocytopenia (<1500). Anecdotally, progressive lymphopenia corresponds with worsening lung disease with the theory being increased lymphocyte sequestration in the lung. CRP: Elevation associated with infection and progression with worsening disease. For afebrile patients with URI, a number of clinics in China have used the presence of lymphopenia or CRP elevation to dictate further escalation to CT scan and viral testing. Troponin: Elevation associated with progression to severe disease. In the Chinese population there have been many cardiac complications including myocarditis, although anecdotally, this has not been as prevalent in the Italian population.

Pearl: Consider sending procalcitonin, CBC with differential, CRP, and troponin to help support your clinical diagnosis of COVID-19.

Evidence: Mechanism of ACE/ARB: The spike proteins that make the corona on the SARS-CoV-2 bind strongly to human ACE2. 83% of ACE-2 expressing cells in the human are in alveolar epithelial cells in the lung. Expression of ACE2 is increased by ACEis, ARBs, thiazolidinediones, and ibuprofen. Thus, some have postulated that this mechanism may lead to more severe disease. However, this is purely hypothesis-generating and others have even argued ARBs could be protective. Mechanism of Hydroxychloroquine/chloroquine: Inhibits SARS-CoV2 in vitro. Although chloroquine has been shown to inhibit other viruses in vitro, it has failed to be effective in in vivo trials for other viruses. Tiny 6-patient survey in France referenced below suggests decreased viral load on chloroquine + azithromycin. Mechanism Remdesivir: RNA dependent RNA polymerase inhibitor intended to treat Ebola that has shown in vitro activity against MERS, SARS-CoV-1, and SARS-CoV-2. Case report in NEJM suggested improvement and anecdotal reports suggest improvement when used for compassionate use.

Pearl: Both the ESC and AHA recommend continuing ACEs and ARBs as they are medications with proven benefits in stroke, heart failure, HTN, and MI patients and there is no proven harm. This is sound advice for outpatients. However, for an inpatient with coronavirus it is reasonable to consider alternative agents for BP control while remembering that arterial hypertension has been associated with death in China. Because HCQ is cheap, safe and easy to acquire, we have been using this at VNC, although questionable benefit. Clinical trials ongoing, but our ID service has applied for compassionate use in our vented patients. Remember the extensive exclusion criteria (no multi-organ failure, no vasopressors, ALT>5x ULN, CrCl<30, CVVH/HD, pregnancy, or use of other experimental antiviral agents [think lopinavir/ritonavir]).

Evidence: Antihypertensive regimens are often increased at discharge due to elevated blood pressures during inpatient hospitalization. Medication confusion and errors are common at discharge and during this high risk time medication changes may increase risk of adverse drug events and readmissions in older adults. This retrospective, propensity-matched cohort study showed that intensification of antihypertensive regimens at discharge in patients > 65, who were admitted for non-cardiac conditions, was associated with increased rates of readmission (NNH = 27) and medication related adverse events (NNH = 63). It also showed that intensification was not associated with longer term improvement in blood pressure or reduction in cardiovascular events during 1 year follow up.

Pearl: Avoid increasing anti-hypertensive medications at the time of discharge in patients admitted with non-cardiac conditions, particularly in patients with previously well-controlled blood pressure due to increase in medication related adverse events and readmissions without evidence of improvement in long-term blood pressure control.

Evidence: Delirium has been associated with increased length of stay, institutionalization, long-term cognitive impairment and overall mortality. A recent systematic review of randomized controlled trials and prospective observational studies examined the benefits and adverse effects of haloperidol and 2nd generation antipsychotics (ie olanzapine, risperidone, and quetiapine) when compared with placebo for the treatment of delirium in adults. Study selection excluded studies that did not use a validated instrument to diagnose delirium. The review focused on 5 “critical outcomes” which included cognitive function, hospital length of stay, delirium severity, sedation, inappropriate continuation of antipsychotics, 2 clinical outcomes which included delirium duration and mortality, and 2 safety outcomes which included cardiac and neurological mortality. The patient populations from the various studies included noncritically ill and critically ill patients as well as hospice/palliative patients. There was no difference between haloperidol and 2nd generation antipsychotics vs placebo in sedation status, delirium duration, hospital length of stay, or mortality. There was no difference in delirium severity and cognitive functioning for haloperidol versus second-generation antipsychotics, with insufficient or no evidence for antipsychotics versus placebo. There was little evidence demonstrating neurologic harms associated with short-term use of antipsychotics for treating delirium in adult inpatients, however there were potentially harmful cardiac effects particularly prolongation of the QT interval with second-generation antipsychotics versus placebo or haloperidol.

Pearl: Current evidence does not support routine use of haloperidol or second-generation antipsychotics to treat delirium in adult inpatients, which is aligned with guidelines from the Society of Critical Care Medicine (2018).

Evidence: Anticoagulation for VTE associated with cancer has posed a significant management dilemma for clinicians. Anticoagulation is an effective prophylactic and treatment for VTE but increases risk for major bleeding, which can be equally morbid and life-threatening. The risks of VTE, and the harms and benefits of anticoagulation, are finely balanced and differ for each patient depending on their cancer, comorbidities, and treatments. The ASCO guidelines incorporate new data from multiple RCTs and meta-analyses that assessed direct oral anticoagulants (DOACs) vs placebo in some settings, and DOACs vs low-molecular-weight heparins (LMWH), unfractionated heparins, and vitamin K antagonists in other settings.

Pearl: Evidence now supports the use of rivaroxaban for the treatment of VTE associated with malignancy; however, rivaroxaban is linked with a higher risk of bleeding compared to LMWH in patients with GI and genitourinary cancers. Consider use of rivaroxaban in non-GI and non-genitourinary cancers. Consider the recommendations below when managing a patient with active malignancy. The Chest guidelines for VTE prevention and treatment have not yet been updated with this latest evidence. Revised Recommendations (Strength of Recommendation): 1.High-risk outpatients with Khorana score ≥ 2 may be offered thromboprophylaxis (apixaban, rivaroxaban, or LMWH) when starting new systemic chemotherapy if they have no significant risk factors for bleeding and no drug interactions. (Moderate) 2.Pharmacologic thromboprophylaxis for patients having major surgery for cancer should be continued for at least 7 to 10 days. Extended prophylaxis with LMWH for up to 4 weeks after surgery is recommended for patients having major open or laparoscopic abdominal or pelvic surgery for cancer who have high-risk features. (Moderate to strong) 3.In patients with cancer and established VTE, initial anticoagulation may include LMWH, UFH, fondaparinux, or rivaroxaban. Initial parenteral LMWH is preferred over UFH for patients with newly diagnosed VTE and without severe renal impairment (creatinine clearance < 30 mL/min). (Strong)

Evidence: In a multi-center RTC of 2003 adult ICU patients, adding pneumatic compression devices to pharmacological DVT prophylaxis with either unfractionated heparin or low-molecular-weight heparin made no difference in the incidence of lower extremity DVT (3.9% vs 4.2%) or overall VTE events (10.4% vs 9.4%). The study included medical, surgical, and trauma ICU patients with a mean age of 58, who received the treatment for a median of 7 days.

Pearl: Do not order SCDs on ICU patients who are already receiving pharmacological DVT prophylaxis.

Evidence: Recent studies have suggested that a shorter antibiotic course (5-day duration) is safe and efficacious for the treatment of community acquired pneumonia (CAP) in most patient populations. There are risks associated with longer treatment duration, including the development of antibiotic resistance and Clostridium difficile infection. This retrospective cohort study showed that two thirds of patients receive excess duration of antibiotics for CAP (median 8 days) and that the majority of excess duration was prescribed at the time of discharge. Patients who were more likely to receive excess duration of antibiotics were those who had additional testing ordered such as respiratory cultures, nonculture diagnostic testing, recent antibiotic use, or treatment duration not documented in the discharge summary. There were no differences in readmission rates, mortality, or C diff in patients receiving excess duration of antibiotics, but there was an increase rate of patient-reported antibiotic associated adverse events.

Pearl: For patients admitted with community acquired pneumonia, consider a 5-day duration of antibiotics and documenting the total duration of antibiotics in the discharge summary.

Evidence: The now outdated 2007 (IDSA/ATS) guideline on community-acquired pneumonia (CAP) recommended urinary antigen testing (UAT) for Streptococcus pneumoniae (SP) and Legionella pneumophila (LP) in patients with certain clinical characteristics. However, these recommendations were based on expert opinion and had not been validated. A prospective cross-sectional study, which consisted of 1971 patients who had inpatient pneumonia hospitalizations from 2010-2012, evaluated the sensitivity and specificity of recommended indications for SP and LP urinary antigen testing from the 2007 (IDSA/ATS) guideline.

Pearl: The new 2019 IDSA/ATS pneumonia guideline does not recommend routine urine SP or LP testing but it can be considered in cases of severe pneumonia (conditional recommendation, low quality of evidence). Urine LP testing can also be considered in cases when indicated by epidemiological factors, such as association with a Legionella outbreak or recent travel (conditional recommendation, low quality of evidence).

Evidence: RSV is no longer considered a pediatric disease (laryngotracheobrochitis or “croup”) and can cause exacerbation of cardiorespiratory disease in patients ≥60 years old from the community and chronic-care facilities. Medical records of patients ≥ 60 years of age from Kaiser Permanente Southern California were retrospectively reviewed during five winter seasons between 2011 and 2015 and incidence, comorbidities and sequelae associated with RSV versus Influenza were compared. Patients with RSV tended to have greater rates of exacerbation of heart failure, COPD, and asthma in addition to an increased incidence of pneumonia, admissions to the hospital and ICU, longer length of stay, and greater requirement of long-term care after discharge.

Pearl: There is no currently available vaccine to prevent RSV and no highly effective antiviral therapy (including ribavirin), but it is highly communicable and can increase inpatient morbidity and mortality. If you have a patient with flu-like illness with influenza testing negative, consider sending “respiratory viral limited panel” and initiating appropriate precautions until RSV ruled out.

Evidence: Acute respiratory failure is a diagnosis that we often forget to use – and it matters. When documenting acute respiratory failure, it increases the estimated Length of Stay (LOS), Severity of Illness (SOI), Risk of Mortality (ROM) and reimbursement. It also adds more specificity for patients presenting with a variety of respiratory conditions like pneumonia, pleural effusions, COPD, and pulmonary edema. Hospitalists tend to not document this because they think it only relates to patient intubated in the ICU, but we have to change that mentality. Criteria for diagnosis: pO2 less than 60 mm Hg (hypoxemia); pCO2 greater than 50 mm Hg (hypercapnia) with pH less than 7.35; Signs and symptoms of acute respiratory distress. An ABG might be standard for diagnosis but you can also make the diagnosis without an ABG if a patient has an oxygen saturation less than or equal to 90% is documented with physical exam signs of respiratory distress (tachypnea, nasal flaring, accessory muscle use).

Pearl: One needs to document two of the three criteria to formally diagnose acute respiratory failure: pO2 less than 60 mm Hg (or room air oxygen saturation less than or equal to 90%), pCO2 greater than 50 mm Hg with pH less than 7.35, and signs/symptoms of respiratory distress. Document physical exam findings that correlate with acute respiratory failure (RR less than 20 or greater than 10, wheezing, nasal flaring, accessory muscle use, etc).

Evidence: Patients with gout have increased cardiovascular risk. Febuxostat, a nonpurine xanthine oxidase inhibitor, is more effective at lowering uric acid levels when compared to allopurinol. The Cardiovascular Safety of Febuxostat and Allopurinol in Patients with Gout and Cardiovascular Morbidities (CARES) study, a multi-center, randomized, double-blind, non inferiority trial compared cardiovascular outcomes associated with febuxostat and allopurinol in patients with gout and cardiovascular disease. The study showed in patients with gout and major cardiovascular coexisting conditions, febuxostat was shown to have an increased cardiovascular and all-cause mortality. Following the CARES study, the FDA issued a box warning for febuxostat in 2019.

Pearl: Febuxostat should be reserved for patients who have failed or cannot tolerate allopurinol. When prescribing febuxostat, the associated CV risks should be discussed with the patient. Patients who are on febuxostat should be monitored for cardiovascular signs and symptoms.

Evidence: Tramadol is a SNRI which interferes with pain pathways. It is metabolized to ODT, a mu-1 agonist, like opioids. The conversion from SNRI to ODT is done by CYP2D6, which has a wide variation in function across patient populations, making it difficult to assess how a patient will respond. Population of Ethiopians and Saudia Arabians are rapid metabolizers via CYP2D6 (causing increased opiate affect) while those in Scandinavia appear to have a slower metabolism (causing more SNRI effect). Multiple common medications can block the CYP2D6 enzyme and can even precipitate opioid withdrawal in patients taking tramadol. In patients with CKD, the drug interactions are particularly important as patients on dialysis take an average of 19 pills per day. More concerning, among patients aged 50 years and older with osteoarthritis, tramadol was associated with a significantly higher rate of mortality over 1 year of follow-up compared with commonly prescribed non-steroidal anti-inflammatory drugs.

Pearl: Medications that can block CYP2D6 include bupropion, haldol, fluoxetine, amiodarone, and cinacalcet. Tramadol can cause serotonin syndrome when taken with other serotonergic medications. Other significant risks include increased seizures and hypoglycemia. In patients >50 years old with osteoarthritis, tramadol was associated with a mortality rate than NSAIDs.

Evidence: Researchers followed a cohort of over 990,000 UK patients newly started on anti-hypertensives for an average of 6.4 years and found that use of ACE-I was associated with a modestly increased risk of lung cancer compared with ARBs (1.6 vs 1.2 per 1000 person years, HR 1.14). There is a cumulative effect and the risk increases with duration of use beyond the initial 5 years (HR up to 1.31 for >10 years of use). The possible explanation for this association is that ACE-I causes accumulation of bradykinin in the lung, which has been reported to stimulate lung cancer growth. ACE-I also increases substance P, which is expressed in lung cancer tissue and linked to tumor proliferation and angiogenesis.

Pearl: The link between long term ACE-I use and increased risk of lung cancer seems to be a plausible one. Keep this association in mind when starting patients on new anti-hypertensives, especially in young patients. Educate patients regarding the risks and benefits ACE-I therapy.

Evidence: Because the risk of these serious side effects generally outweighs the benefits for selected patients, the FDA has determined that fluoroquinolones should be reserved for use in patients with serious bacterial infections, including anthrax, plague and bacterial pneumonia among others.

Pearl: Side Effects Include: Increased risk of tendinitis and tendon rupture, Increased risk of C-Diff Colitis, Potential for irreversible peripheral neuropathy, Increased risk of aortic dissections or rupture of an aortic aneurysm, Increased risk of fatal hypoglycemia occurring more frequently in the elderly and those taking oral hypoglycemics or insulin, Increased risk of disorientation, agitation, and delirium

Evidence: Colonic diverticulitis is an inflammatory process that most commonly affects the sigmoid colon. Diverticulitis is inflammation of diverticula and can be either uncomplicated or complicated (i.e., associated with abscess, fistula, stricture, or perforation and peritonitis). Rates of diverticulitis are increasing in association with increasing rates of obesity. Most cases are straightforward for hospitalists to manage but there are helpful clinical pearls surrounding outpatient management and patient education that are highlighted in this article.

Pearl: Simple diverticulitis in the absence of high fever, clinically significant laboratory or radiologic abnormalities, or immunosuppression, can be managed on an outpatient basis. The severity of recurrent episodes of diverticulitis is generally similar to that of the initial episode. Symptoms typically improve within 2-3 days after the initiation of treatment, at which time the diet is commonly advanced to clear liquids and then to a low-residue diet, although data from randomized trials to guide inpatient dietary management is limited. Consultation from nutrition is recommended to supplement physician education. So what the heck is a low residue diet anyways? Limiting high fiber foods (in the initial 1-2 weeks after episode of diverticulitis). 'Residue' is undigested food, including fiber, that makes up stool. The goal is to have fewer, smaller BMs each day. Risk factors for diverticulitis include smoking, the use of nonsteroidal anti-inflammatory drugs, physical inactivity, obesity, diets low in fiber, diets high in refined carbohydrates, and red meat. Evidence does not support the idea that seeds, nuts, and popcorn cause diverticulitis.

Evidence: Multi centric randomized controlled five year observational follow up of 530 patients (ages 18-60) with CT confirmed uncomplicated acute appendicitis, compared appendectomy versus antibiotic therapy. At 5 years the surgical group had a significantly higher overall complication rate (surgical site infections, incisional hernia, abdominal pain) compared to the non-surgical group. About 39% (1/3) of the non-surgical or antibiotic group developed recurrent appendicitis needing surgery, but no complications in the antibiotic group were related to delay in surgery.

Pearl: Long term follow up indicates that antibiotics alone could be a reasonable alternative to surgery in patients with uncomplicated appendicitis. This data provides hospitalists evidence to present to patients and families when reviewing management options. In complicated appendicitis (perforation, appendolith) and in older patients, surgery may still be the best option.

Evidence: A European multi-center RTC of 3298 ICU patients with at least one risk factor for GI bleeding (ie shock, anticoagulation, dialysis, mechanical ventilation, liver disease or coagulopathy) showed there was no difference in 90-day mortality (primary outcome) or incidence of clinically significant GIB, pneumonia, C diff infection or MI (secondary composite outcome) for those initiated on PPI prophylaxis vs placebo. There was a reduced incidence of clinically important GI bleeding (NNT = 59) without causing more infections, however the trial was not powered to address each component of the composite secondary outcome, thus limiting conclusions.

Pearl: The short-term benefits of decreased GI bleeding with PPI use in select ICU patients may outweigh any increased risk of adverse outcomes. Longer term clinical benefit, however, has not been demonstrated and would recommend discontinuing the PPI once leaving the ICU.

Evidence: At the American College of Gastroenterology Conference 2018, Dr. Melissa Latorre, Director of Inpatient Services for Gastroenterology at NYU, gave an expert talk on her approach to inpatient constipation. She states that “ineffective laxatives, such as docusate sodium, continue to be the most commonly prescribed.” Since the data is very limited, her experience-based approach to inpatient constipation is like chemotherapy with an induction and maintenance regimen. She warns “don’t be fooled by diarrhea,” as it may be overflow, and that it is important to educate nurses and patients that diarrhea is desired upfront.

Pearl: From above: Induction phase: Osmotic laxative like polyethylene glycol 17 g in 8 oz water BID or TID or colonoscopy prep (movi-prep) if patients can tolerate. AVOID fiber, senna, and lactulose upfront as they can exacerbate the symptoms of pain, bloating and distension. Maintenance: Start fiber, increase free water intake, and oral laxatives. Pro-tip: If available – opiate-receptor antagonists are helpful in patients regularly taking opioids. From below: Induction phase: Manual maneuvers and suppositories. Glycerin helps soften stool and bisacodyl helps with rectal motility. After the first BM, provide enemas (mineral oil or tap water) to clean more proximally. Maintenance: Once BMs become reliable, pull back on rectal therapy.

Evidence: Docusate remains widely prescribed despite multiple RCTs that have shown it has no benefit over placebo. Docusate is a detergent that tastes like soap and may decrease oral intake. It delays the initiation of actually effective laxatives and increases pill burden. Effective alternatives supported by evidence include Miralax, Senna, Psyllium, and Lactulose.

Pearl: Stop using docusate! Continue to educate Surgeons and Surgery APNs/PAs to stop ordering docusate.

Evidence: The incidence of symptomatic DVT and PE in hospitalized patients is 0.96% and 1.2% and asymptomatic DVT is 1.8%. Studies have shown that IPC reduces the risk for venous thromboembolism in high risk patients including orthopedic, surgical, trauma and stroke patients. The largest systematic meta-analysis of 70 studies and over 16,000 high risk patients concluded reduction in the rates of DVT from 16.7% to 7.3% and PE from 2.8% to 1.2%. However, there is no good data showing efficacy of IPCs in general medical ward patients. IPCs may not be as beneficial for medical ward patients because they are frequently worn or applied improperly, can be a fall risk, and can cause skin irritation and breakdown. Different IPC models have varied efficacy and are expensive.

Pearl: Given the costs, unproven efficacy and possible side effects of IPC in general medicine patients, IPC should not be ordered ordered on general medicine ward patients at low or moderate risk. Combined IPC and pharmacologic prophylaxis should only be used for high risk trauma and surgical patients. IPC alone should ordered only for high risk patients who have a contraindication to pharmacologic prophylaxis. Risk stratification can be assessed by the Three Bucket Model or Padua Prediction Score below.

Evidence: The AHA developed guidelines in 2004, updated in 2017, for indications to order and continue telemetry monitoring in hospitalized patients. The Society of Hospital Medicine’s Choosing Wisely campaign has highlighted the importance of discontinuing telemetry for non-ICU patients. Despite these guidelines, inappropriate telemetry use remains high, with up to 43% of patients on telemetry without a clear indication for use. Over-utilization of this resource can result in unnecessary testing, “alarm fatigue” and increase hospitalization costs. It is important to be familiar with these indications and re-assess needs on a daily basis.

Pearl: Inappropriate telemetry use remains high and is likely to increase cost of care and produce false positives potentially resulting in errors in patient management. Reconsider ordering telemetry in patients without a clear indication and re-assess the need for continued monitoring daily.

Evidence: According to two single center trials, one looking at patients admitted to the ICU and the other at patients admitted to the floor, administering balanced crystalloids (LR) in the ED may be superior to normal saline, resulting in a 1% lower rate of a composite measure of 30 day all-cause-mortality, new renal-replacement therapy, or persistent renal dysfunction with the benefit driven by kidney injury in floor patients. Normal saline can cause iatrogenic hyperchloremic metabolic acidosis which can lead to renal vasoconstriction, relative hypotension, and relative hyperkalemia. These physiologic realities have now been shown to have clinical consequences.

Pearl: Favor using LR over NS when giving IVF resuscitation to prevent acute kidney injury, particularly when using large volumes, treating severe sepsis, or treating patients with diarrhea.

Evidence: There is a common culture among clinicians and nurses to consider supplemental oxygen as beneficial in “sick patients” whose oxygenation saturation is in the normal range. In both human and animal studies, hyperoxia can lead to iatrogenic harm and most importantly, higher oxygen levels have been linked to increased mortality risk. Due to these risks, BMJ recently made oxygen therapy recommendations in acutely ill medical patients. Of note, the benefits or harm of hyperoxia have yet to be established in the setting of cerebral ischemia, out of hospital cardiac arrest, surgical settings and cardiac surgery.

Pearl: In acutely ill medical patients on oxygen therapy, keep peripheral oxygen saturation ≤96%. A target oxygen saturation range of 90-94% appears to be reasonable for most patients. Patients at risk of developing hypercapnic respiratory failure, goal oxygen saturation range is between 88-92%. Patients who have carbon monoxide poisoning, cluster headaches, sickle cell crisis, and pneumothorax, high oxygen saturations approaching 100% are still recommended.

Evidence: Multiple studies have evaluated the effectiveness, bleeding risk, and cost/benefits of LMWH vs UFH for DVT prophylaxis

Pearl: Consider LMWH when starting DVT prophylaxis. Avoid LMWH in patients with creatinine clearance < 30 ml/hr. Information on optimal dosing in morbidly obese patients is limited.

Evidence: Four major randomized placebo-controlled trials involving thousands of patients followed over several years show that patients on SGLT2i have a lower risk CV events. The most prominent cardiac risk reduction was for hospitalization for heart failure. This drug class also shows secondary benefit in reducing risk for worsening renal function, end stage renal disease or death due to renal disease. However, studies also show an increase incidence of lower limb amputation and DKA associated with SGLT2i. There are four RCTs (phase 3 clinical trials) evaluating SGLT2i in patients with HFpEF and HFrEF without diabetes. Anticipate that SGLT2i may morph into heart failure drugs rather than primary diabetic agents!

Pearl: Greatest CV benefit for SGLT2i is for hospitalization related to heart failure. Be aware of adverse consequences of lower limb amputation and DKA.

Evidence: For decades, the FDA recommended against the use of metformin in men and women with a serum creatinine > 1.5 and 1.4 respectively. In 2016, the FDA revised its recommendation and instead of an absolute creatinine cutoff, eGFR is utilized to guide metformin prescribing in patients with mild to moderate renal impairment. Now metformin is contraindicated in patients with eGFR < 30 mL/min/1.73 m2. (See table below for more specific guidelines regarding metformin use)

Pearl: Instead of an absolute creatinine cut off, eGFR should be used to determine the discontinuation or continuation of metformin. In patients with progressive liver disease, decompensated CHF, and alcoholism, metformin should be discontinued.

Evidence: The ADA guidelines recommend a Hgb A1c goal 7.0% for most adult patients to reduce the risk of development and progression of microvascular complications. ACP recommends a goal Hgb A1c between 7-8% for most patients with type 2 diabetes. Multiple studies have suggested that there is no significant reduction in cardiovascular disease outcomes and there is an increased risk of severe hypoglycemia in patients with intensive glycemic control who have advanced type 2 diabetes. In patients with limited life expectancy, advanced microvascular or macrovascular complications, extensive comorbidities, a history of severe hypoglycemia, or in patients who it is difficult to achieve target glucose goal despite adequate diabetes education and treatment, less strict hemoglobin A1c goals (< 8%) can be considered.

Pearl: When educating patients and families about blood glucose management based on their A1c, consider an A1c goal (< 8%) for patients with advanced age, more advanced disease and significant comorbidities as risks of strict glycemic control outweighs the benefits.

Evidence: The ADA and AACE guidelines for patients with hyperglycemia who require insulin recommend against the prolonged use of SSI as monotherapy and support the use of basal plus correctional insulin with the addition of nutritional insulin for patients with consistent oral intake. In type 2 diabetics, consider insulin based on a total daily dose of 0.4 units/kg for patients presenting with blood sugar levels ≤200 mg/dL and 0.5 units/kg for those with higher initial glucose levels. Half of the total daily dose can be given as basal insulin, and the other half can be divided among meals. Be cautious with insulin dosing in patients aged &gt;70 years, in those with impaired renal function, and in situations in which steroid doses are fluctuating.

Pearl: Stop prolonged SSI monotherapy into a patient’s hospitalization. Instead, consider adding basal and prandial insulin, using a weight-based approach for insulin-naive patients.

Evidence: 500 episodes of bacteremia were studied at a tertiary hospital during 2015. 140 patients were found to have Gram-negative rod (GNB) bacteremia. Of these, 52 (37%) had follow up blood cultures which were positive in eight patients (15%), including five with Escherichia coli and one each with K pneumoniae, Serratia marcescens, and Stenotrophomonas maltophilia. The repeat blood culture was more likely to be positive in febrile patients. GNB bacteremia is typically transient and usually resolves rapidly after the initiation of appropriate antibiotic therapy and source control.

Pearl: Follow up blood cultures may be unnecessary for gram negative rod bacteremia unless the patient is febrile.

Evidence: A multi-center RCT among patients with ESBL (Ceftriaxone resistant but Zosyn sensitive) E coli or K pneumoniae blood stream infections, definitive treatment with Zosyn compared to Meropenem did not result in a noninferior 30-day mortality.

Pearl: Do NOT use Zosyn for ESBL blood stream infections even if susceptibilities show sensitivity to Zosyn

Evidence: A multi-center open labeled RCT showed that patients hospitalized with gram-negative bacteremia achieving clinical stability (afebrile and hemodynamically stable for 48 hours) before day 7, an antibiotic course of 7 days was non-inferior to 14 days. Reducing antibiotic treatment for uncomplicated gram-negative bacteremia to 7 days is an important antibiotic stewardship intervention. This study mainly looked at Enterobacteria from urinary source.

Pearl: You may consider a 7 instead of 14-day course of antibiotics for patients with clinically stable gram-negative bacteremia (think UTI with sepsis).

Evidence: A UK RTC showed that oral antibiotics were non-inferior to IV antibiotics when used for the first 6 weeks for osteomyelitis and septic arthritis and resulted in similar treatment failure rates at 1 year.

Pearl: You may see ID recommending highly bioavailable oral antibiotics for adherent patients with osteomyelitis or septic arthritis in the future, so hold off on automatically ordering a PICC.