The Unit: A Cardiology Blog

Sunday, September 16, 2012

PILL-CVD: Clinical services and keeping ourselves honest

Given the increasing cost and complexity associated with health care delivery, a variety of stakeholders from across the continuum of care (e.g., hospitals, health systems, third party payers, state and federal government agencies) have launched exhaustive efforts to identify strategies that improve outcomes as patients transition from the health care environment to home. Challenges related to medication use -- from cost to compliance to adverse drug events -- serve as a key area of emphasis, especially given their contribution to patient injury and hospital readmissions. For example, adverse events related to medication therapy have been associated with patient injury in up to 40% of hospitalized patients and nearly 20% of patients following discharge [1,2]. Given the expertise of pharmacists in the area of medication safety, many see us as being the health care professionals responsible for leading efforts aimed at overcoming these challenges. A number of initiatives have been investigated as a result, including medication reconciliation at admission and/or discharge, discharge counseling and education, telephone follow-up, and more.

Recently, the PILL-CVD trial [3] was designed to evaluate whether a combination of these strategies would reduce the number of clinically important medication errors in patients being discharged after an acute coronary syndrome or heart failure exacerbation. Patients (n = 851) were randomized to usual care or a pharmacist-based intervention that included discharge medication reconciliation, discharge counseling, and individualized telephone follow-up after discharge. Although clinically important medication errors were discovered in over half of the patients enrolled in the study, intervention by a pharmacist failed to reduce them when compared to usual care. Patients in the intervention group tended to have fewer potential adverse drug events, but even this difference was not statistically significant.

There were certainly limitations of the study, many of which were discussed by the authors. For example, patients were enrolled at two large academic medical centers, where robust medication reconciliation systems may have made it difficult to demonstrate a difference in the intervention group. Additionally, the patient population was relatively well-educated and demonstrated a high level of health literacy, which may have also impacted outcomes.

While the results of PILL-CVD may have rustled some feathers, I can accept it because I think it helps keep us honest. Our clinical activities should be just as evidence-based as the therapies we recommend, especially in light of the rising costs of care and the already limited resources in pharmacy and beyond. We should support, extend, and improve those activities that optimize patient outcomes and abandon those that do not. Where some may see the results of PILL-CVD as a threat to the role of pharmacists in transitions of care, I instead see it as a call to re-focus on those initiatives that do improve medication therapy outcomes, some of which were already included in this study.

As one example, pharmacists in the acute care setting are ideally positioned to perform comprehensive medication reconciliation at discharge. To clarify, I share the view that discharge medication reconciliation is not simply the generation of a medication list, but is instead a vital component of medication therapy management -- that is, an interdisciplinary strategy aimed at optimizing the medication regimen of each individual patient -- an effort that has been associated with improvements in both clinical and economic outcomes in a variety of settings [4]. As such, medications should not be evaluated solely on the basis of whether or not the patient was receiving them during their hospitalization, but whether it is appropriate to continue them based on factors that include safety, efficacy, cost, and compliance.

On the other hand, I maintain that medication counseling performed by acute care pharmacists at the point of discharge is not a very effective -- nor efficient/sustainable -- strategy for improving medication-related outcomes. While small studies have shown some improvements compared to usual care, differences are marginal at best and other studies have shown no impact at all. When I provide discharge medication counseling, I often see individuals and families who are exhausted, buried in paperwork, and overwhelmed with information... but even more importantly, I see people who are eager to leave the hospital. The last thing they want is for yet another health care professional to provide them with even more information they will soon forget -- and understandably so -- in their hurry to get things together and prepare for discharge. So I reluctantly hand them another stack of handouts to file away in a folder along with their discharge summary, upcoming appointments, laboratory results, medication list, etc.

While this may seem discouraging, it just reminds me to remain focused on those activities that do improve patient outcomes, including:

Rounding in the intensive care unit [5];
Rounding with a cardiology medicine service [6];
Collaborative management of anticoagulation [7], blood pressure [8], dyslipidemia [9], and other cardiovascular risk factors [10]; and,
Improving medication adherence among high-risk patients [11].

So while the results of PILL-CVD may have been concerning for some, I am encouraged that it will lead us to explore new or alternative strategies for improving medication-related outcomes as patients make the transition to the ambulatory care environment. Given significantly limited resources across health care settings and providers who are already stretched thin, it is important that we hold ourselves accountable for performing those clinical activities that are best supported by the available evidence -- even if it means leaving behind some activities that we once held dear.

Comments welcomed and encouraged.

References

Monday, September 3, 2012

Update: Deja vu in ATLAS ACS-2: rivaroxaban in the subset of patients with ST-segment elevation myocardial infarction

In an article posted on theheart.org last week highlighting the results of several late-breaking clinical trials revealed at the 2012 European Society of Cardiology (ESC) meeting, experts again debated the utility of adding low-dose rivaroxaban (2.5 mg twice daily) to dual antiplatelet therapy in patients with acute coronary syndromes (ACS). The discussion involved a pre-specified sub-analysis of ATLAS ACS 2 [1], a trial that randomized patients with ACS who were already receiving aspirin and clopidogrel to either rivaroxaban or placebo. A sub-analysis of the patients with ST-segment elevation myocardial infarction (STEMI) found a small but marginally significant reduction in major adverse cardiovascular events (absolute risk reduction 1.9%, HR 0.81 (95% CI 0.65-1.00), p = 0.047) with rivaroxaban. However, as observed in the overall cohort of patients with ACS, this benefit was accompanied by a significant increase in several types of bleeding (details unavailable at this time).

As I describe in a previous entry on the results of ATLAS ACS 2, I am still not sure it is worth pursuing the addition of rivaroxaban in patients with ACS until we know its risk versus benefit profile in combination with aspirin and one of the newer P2Y12 inhibitors prasugrel or ticagrelor. Although, given the results of another trial revealed at ESC 2012, perhaps we should be dropping aspirin from the "triple therapy" strategy anyway.

What is perhaps an even more challenging scenario, as the article goes on to discuss, is what to do when patients have multiple indications for anticoagulation or are already receiving rivaroxaban for a different indication and have an ACS event. The appropriate dose of rivaroxaban depends on indication -- 10 mg once daily for prophylaxis of venous thromboembolism following major orthopedic surgery, 15 mg twice daily for the treatment of deep vein thrombosis or pulmonary embolism (assuming the agent will be approved for these indications based on the results of the two EINSTEIN investigations [2,3]), or 20 mg daily for atrial fibrillation (15 mg once daily if impaired renal function). If someone has an ACS event that requires dual antiplatelet therapy and is already taking rivaroxaban for one of these indications, do we reduce the dose to 2.5 mg twice daily given that higher doses are already known to increase the risk of bleeding in this population [4]?

I suppose nobody has any definitive answers to these and other challenging clinical scenarios that will soon be at our doorstep, but one thing I do know for sure -- what an exciting time to be practicing in the area of cardiology.

Note: the 2.5 mg twice daily dose of rivaroxaban was up for approval by the US Food & Drug Administration earlier this year, but failed to gain the indication for use in acute coronary syndromes.

References

Tuesday, August 21, 2012

Going to ground: risk of falls and oral anticoagulation

LAW #2: GOMERS GO TO GROUND.

In The House of God, a 1970s novel detailing his experience as a first-year medical resident, Samuel Shem (a pseudonym for Stephen Bergman, MD), cites the above axiom as one of the Laws of the House of God taught to him by a senior resident known only as The Fat Man. The rule itself refers to the propensity of elderly patients to fall or fall from their hospital beds, and like many concepts from the novel that have made their way into the medical vernacular, an assessment of fall risk remains an important evaluation in both hospitalized and ambulatory patients and often significantly impacts the strategies of care offered or delivered to them.

One of the most common scenarios during which a discussion of fall risk occurs is when determining the appropriate antithrombotic strategy for an elderly patient with atrial fibrillation (assuming their annual risk of stroke warrants oral anticoagulation, as determined by the CHADS₂ and/or CHA₂DS₂-VASc score). The risk of major hemorrhage associated with a traumatic fall is often cited as a reason not to provide full anticoagulation, but evidence from the literature indicates these risks are often overestimated; more importantly, the preference for aspirin in many of these patients is unlikely to ameliorate these risks.

For example, in the Birmingham Atrial Fibrillation Treatment of the Aged Study (BAFTA) [1], patients over the age of 75 years (mean 81.5 years) with atrial fibrillation were randomized to aspirin 75 mg daily or adjusted-dose warfarin (INR goal 2-3); after a mean follow-up period of nearly 3 years, adjusted-dose warfarin was associated with a reduction in the number of total strokes compared to aspirin (3.4% versus 1.6%, respectively, p = 0.003), a difference that was largely driven by a difference in ischemic strokes (2.5% versus 0.8%, p = 0.0004). More importantly, the benefits observed with warfarin were not accompanied by an increase in major hemorrhage or hemorrhagic stroke. In another analysis [2], the risk of falls among elderly patients with atrial fibrillation was not found to be a major contributor to the hemorrhagic events associated with warfarin -- in fact, a patient would need to fall nearly 300 times in a single year for the risks to outweigh the potential benefits.

Given the alternatives now afforded by the new oral anticoagulants apixaban, dabigatran, and rivaroxaban -- which are arguably safer than warfarin in appropriately selected patients -- the benefits associated with full anticoagulation are even more likely to outweigh the hemorrhagic risks commonly attributed to falls. In fact, when it comes to intrancranial hemorrhage, the most devastating complication of a traumatic fall, each of the new oral anticoagulants is associated with a markedly lower risk compared to warfarin -- a difference in absolute risk per year of 0.24%, 0.44%, and 0.20% with apixaban, dabigatran, and rivaroxaban, respectively [3-5].

Thus, as the population continues to age and the incidence of atrial fibrillation grows, it is likely that discussions involving the appropriate antithrombotic strategy in elderly patients with this condition will only become more frequent as time goes on. While every patient should be evaluated for their individual risk of stroke and hemorrhage, we should be cautious not to potentially deny them the benefits of oral anticoagulation in favor of an antithrombotic strategy (i.e., aspirin) that is definitively less effective but also unlikely to ameliorate the risk of hemorrhage associated with a traumatic fall.

References

Friday, August 3, 2012

Shopping around: variations in price for generic clopidogrel

In May of this year, the patent for brand-name clopidogrel (Plavix^®) finally expired, an event that was heralded as a breakthrough in the management of coronary artery disease. As with all newly available generic medications, months may pass before generic clopidogrel becomes considerably less expensive than its brand name counterpart. As generic manufacturers scramble to ramp up production, significant differences in price and availability are likely.

These differences were recently highlighted in the September issue of Consumer Reports, where a variety of pharmacies across the US were surveyed on the price of a 30-day supply of clopidogrel. The prices ranged anywhere from $15 to $200 -- notably, independent pharmacies consistently quoted a price of around $50, which was often markedly lower than many of their chain and supermarket competitors.

The article also mentions the Plavix^® Choice program, which allows patients to obtain brand-name Plavix^® through the manufacturer at a significantly reduced cost -- $37 in most cases. Significant restrictions apply, of course; for example, only commercially-insured and cash-paying patients (i.e., no Medicare or Medicaid enrollees) are eligible for the program. Although the article goes on to claim that the manufacturer picks up the difference, it is not clear from the Bristol-Myers Squibb website that this is the case. When the patent for brand-name atorvastatin (Lipitor^®) expired in the fall of last year, Pfizer was widely criticized for several controversial financial arrangements with pharmacy benefit managers (PBMs), including a prescription discount program aimed at retaining market share. Although the program provided some cost-savings for patients, insurers and health-systems were often stuck paying the bill.

In the end, the editors of Consumer Reports recommend that patients "shop around" for generic clopidogrel. However, as most health care professionals will say, shopping around is rarely the right answer in the long run. Patients should obtain all of their medications at a single community pharmacy (or pharmacy chain/franchise), as this allows the pharmacist to maintain a complete and accurate medication record, which is critical for evaluating drug-food and drug-drug interactions (see here for an example), responding to requests for medication information from other health care providers, assisting in medication reconciliation when patients are hospitalized, and more. Besides, most pharmacies already offer competitor price-matching programs, so no one should need to obtain their clopidogrel prescription anywhere other than their home pharmacy.

So while it is appropriate to recommend that patients ask around for the lowest price of generic clopidogrel, they should still be encouraged to shop in only one place.

Sunday, July 22, 2012

Bridging with enoxaparin in patients with mechanical heart valves

We often have patients with mechanical heart valves who require temporary discontinuation of warfarin in preparation for surgery or an invasive procedure. Because these patients are at such high risk for valve thrombosis, they are commonly bridged with a parenteral anticoagulant until the INR becomes sub-therapeutic. Prior to a procedure, intravenous (IV) unfractionated heparin (UFH) is usually preferred, as the agent can be discontinued several hours prior to the procedure. Immediately afterwards, patients require bridging back to warfarin until their INR is therapeutic again. I usually recommend enoxaparin, a low molecular weight heparin (LMWH) for this purpose, as it allows patients to be discharged and have their INR managed as an outpatient. I sometimes get surprised looks from the team, as many clinicians feel compelled to keep patients hospitalized on a UFH infusion until their INR is therapeutic.

I can understand their reluctance given the "Dear Health Care Professionals" letter issued in 2002 by Sanofi-Aventis, the manufacturer of Lovenox (enoxaparin). The letter accompanied a labeling change where providers were urged not to use enoxaparin in patients with mechanical heart valves. Although the labeling revision stopped short of making it a black box warning, many clinicians considered the letter to essentially mean just that.

The labeling revision was prompted by a small study in pregnant women with mechanical heart valves in which enoxaparin was used as the sole anticoagulant (i.e., not as a bridge to warfarin). Given the teratogenicity of warfarin, an alternative method of anticoagulation is necessary to prevent valve thrombosis in these patients during pregnancy. Unfortunately, the trial was terminated early due to an increased rate of valve thrombosis and death. Although pregnant women are already at an increased risk for thromboembolism and have altered pharmacokinetics (e.g., volume of distribution) that may cause them to respond differently to enoxaparin than non-pregnant patients, the risk associated with enoxaparin was extrapolated as applying to all patients with mechanical heart valves.

However, in the years following the 2002 revision, the approved label has since been softened to state the following:

The use of enoxaparin has not been adequately studied for thromboprophylaxis in patients with mechanical prosthetic heart valves and has not been adequately studied for long-term use in this patient population. Isolated cases of prosthetic heart valve thrombosis have been reported in patients with mechanical prosthetic heart valves who have received enoxaparin for thromboprophylaxis.

Isolated cases of valve thrombosis have also been reported with subcutaneous UFH, intravenous UFH, and warfarin (even in the setting of a therapeutic INR). In other words, patients with mechanical heart valves are at high risk for thromboembolic events, no matter the anticoagulation strategy utilized.

So what are clinicians to do?

Evidence exists to support the use of enoxaparin as a bridge to warfarin in non-pregnant patients with mechanical heart valves [1]. In a prospective study of 250 patients being transferred to a rehabilitation center following mechanical valve replacement surgery, no increased rates of valve thrombosis or bleeding were observed with the use of an enoxaparin bridge to warfarin. Patients had been managed with an IV UFH infusion prior to transfer to the rehabilitation facility, where they continued enoxaparin until their INR was therapeutic. The transition from UFH to enoxaparin occured at approximately 16 ± 11 days after surgery, but the mean INR was subtherapeutic (1.5 ± 0.3) at this time of transition. Furthermore, the mean duration of enoxaparin therapy was 8.3 ± 6.0 days. Given the high risk for thromboembolic events present in the population enrolled (e.g., advanced age, concomitant risk factors), one would have expected thrombotic events to occur had the therapy been ineffective.

The practice of using enoxaparin is also supported by the 2012 CHEST guidelines [2], although they give it their lowest recommendation and evidence rating (Grade 2C). That being said, the guidelines do not recommend any alternatives that have better evidence ratings -- in fact, they even recognize prophylactic doses of UFH and LMWH as being better than IV UFH, although I would never recommend these strategies in such a high-risk population

In summary, I tend to recommend the use of therapeutic enoxaparin as a bridge to warfarin in patients with mechanical heart valves after surgery or invasive procedures. The majority of these patients already have a stable warfarin regimen and should obtain a therapeutic INR within a short period of time. Besides, even if IV UFH was marginally better than LMWH, this advantage would probably be outweighed by the adverse effects of prolonged UFH infusions (e.g., heparin-induced thrombocytopenia) as well as the need to remain hospitalized for extended periods of time, where the risk of thromboembolism (not to mention infection, falls, and a myriad of other complications) are also higher. Given the current evidence to support this practice, as well as recognition by clinical practice guidelines, providers should feel comfortable discharging patients on an enoxaparin bridge to warfarin and having their INR followed in the ambulatory setting.

References

Saturday, July 21, 2012

Just a little heart attack: supporting the Go Red for Women campaign

During her review of the initial management of acute coronary syndromes in my Advanced Cardiac Life Support (ACLS) course this afternoon, Paula Miller shared with us the video posted below, produced by the American Heart Association as part of its Go Red for Women campaign. The goal of Go Red for Women is to raise awareness of cardiovascular disease as the number one cause of death among women -- more deaths than all types of cancer combined. As part of the campaign, women are encouraged to understand their risk of cardiovascular disease and recognize the signs and symptoms of heart attack and stroke, which often present differently in women than in men.

Although chest pain or pressure remains the most common symptom of heart attack, women are more likely to experience other symptoms, including:

Lightheadedness
Shoulder or jaw pain
Nausea or vomiting
Shortness of breath
Breaking out in a cold sweat

The signs and symptoms of stroke are generally similar in men and women and include:

Confusion
Dizziness
Loss of balance
Changes in speech
Numbness or tingling in the face, arms, or legs, especially on one side of the body

You can read more about the signs and symptoms of heart attack and stroke here.

Entitled "Just a Little Heart Attack", the video stars Elizabeth Banks as a multitasking mother of two who is busy juggling her career and household while virtually ignoring an ongoing heart attack -- a perfect illustration for why the Go Red for Women campaign is so important. I found it well-done and entertaining and thought I would share:

Sunday, July 8, 2012

Analgesia and sedation in mechanically ventilated patients

Although significant advancements have been made to help guide clinicians in the appropriate management of analgesia and sedation in patients requiring ventilator support, areas of controversy remain. Further complicating the issue is that optimal management strategies may differ based on patient population (e.g., surgical versus medical patients), underlying disease states (e.g., patients with or without underlying pulmonary disease), and concomitant conditions that may affect drug pharmacokinetics (e.g., therapeutic hypothermia, hepatic or renal disease, heart failure). The following is a collection of evidence-based strategies for managing analgesia and sedation in medical patients requiring ventilator support.

Optimizing the patient environment. While most patients will require pharmacologic therapy for the management of agitation, several non-pharmacologic strategies have also been shown to improve outcomes in patients admitted to the intensive care unit (ICU) [1]. Many of these interventions are aimed at maintaining patient sensorium, while others focus on both short- and long-term mental and physical rehabilitation. Non-pharmacologic strategies for optimizing the patient environment include reducing interruptions and ambient noise, promotion of the physiologic sleep-wake cycle, music and recreational therapy [2,3], and early mobilization and physical rehabilitation [4,5].

Minimization of sedative use and preference for opioid analgesics over traditional sedatives (e.g., propofol, benzodiazepines) as the first-line strategy for managing agitation.While agitation is common among patients requiring ventilator support and may complicate ongoing management [6], optimal strategies for managing analgesia and sedation have not been well-defined. Furthermore, uncontrolled discomfort and pain are the most common sources of agitation in critically ill patients, but many are incapable of accurately expressing or describing pain. A common misconception is that amnesia (i.e., not remembering an ICU stay) is helpful in the critically ill population; however, appropriate sensorium and memories have been shown to improve both short-term and long-term psychological outcomes [7]. Additionally, while sedated patients may not recall some details of their hospitalization, survivors of an ICU admission often recall inadequately controlled pain [8].

Although some patients may require the use of sedatives (e.g., propofol, benzodiazepines) for the management of agitation, these agents should be used sparingly, as they are associated with numerous complications in critically ill patients, including gastrointestinal hemorrhage, venous thromboembolism, ventilator-associated pneumonia, bacteremia, and post-traumatic stress disorder (PTSD) among ICU survivors [7,9]. While opioid analgesics are not without adverse effects, they have the advantage of providing both analgesia and sedation without amnestic effects. In fact, when compared to a combination of opioid analgesics and sedatives (i.e., propofol, midazolam), the use of opioids alone has been shown to reduce duration of ventilator support, ICU length of stay, and overall hospital length of stay; the practice may also improve the ability to detect delirium among critically ill patients [10].

Preference for intermittent intravenous boluses over continuous intravenous infusions. Compared to intermittent intravenous boluses, the use of continuous infusion sedatives (including sedative doses of opioid analgesics) in critically ill patients is associated with worsening outcomes, including longer durations of ventilator support, ICU length of stay, and overall hospital length of stay [11,12]. Additionally, patients receiving sedation via continuous infusion are also more likely to be oversedated than those receiving intermittent boluses [13].

The pharmacologic advantages of continuous intravenous infusions (e.g., fentanyl, midazolam) are primarily intended for those patients in whom the duration of mechanical ventilation is expected to be very short (< 24-48 hours, i.e., surgical patients, post-operative cases), where rapid awakening is anticipated and desired. However, the duration of ventilator support is typically longer in critically ill medical patients, due both to the severity of illness and irreversible nature of many of the concomitant disease states that characterize this patient population. As a result, the adverse effects of continuous intravenous infusions (e.g., accumulation of parent drug and active metabolites, tissue distribution, drug-specific infusion syndromes, oversedation) are more likely to outweigh their pharmacologic advantages.

Identification and management of delirium as an adjunct to analgesic and sedative therapy. Although it may be present in up to 80% of patients in the intensive care unit [1], ICU delirium often goes undiagnosed, where it contributes to a number of adverse outcomes, including both short and long-term psychiatric morbidity (e.g., cognitive impairment, functional disability) and mortality [14,15]. The identification and appropriate management of delirium is further complicated by the fact that many agents used for the management of analgesia and sedation not only mask delirium but may also contribute to it. As a result, routine evaluation for the presence of delirium in ICU patients is currently recommended [1]. The most well-validated scale for identifying delirium in critically ill patients is the Confusion Assessment Method for the ICU (CAM-ICU), where a diagnosis of delirium may be made in up to 90% of cases [16].

Neuroleptics are considered the agents of choice for the management of ICU delirium. Although some alternatives exist, haloperidol is the most extensively studied in the setting of agitated delirium. Loading doses may be used for acutely agitated patients while lower doses may be scheduled every 4-6 hours once the acute episode is well-controlled. One of the most concerning adverse effects of haloperidol is its propensity to cause or contribute to QT-interval prolongation. While caution should be exerted with the use of haloperidol in patients with a history of advanced heart disease, QT-interval prolongation is often dose-related, both in terms of cumulative dose (> 50 mg per day) and when high intermittent doses (e.g., > 20 mg) are given intravenously [17]. The risk of torsades de pointes is likely overestimated, as studies and case series suggest an incidence of < 4% [17-18]. While incidence of QT-interval prolongation is reportedly lower with atypical and other newer generation antipsychotics, this assertion is not well-supported by the literature.

Daily interruption of sedative therapy as a strategy to wean pharmacologic therapy. After it was shown to shorten the duration of ventilator support and ICU length of stay in a large randomized controlled trial, performing a daily interruption of sedative therapy (i.e., “daily awakening”) has been advocated as a standard of care for critically ill patients receiving mechanical ventilation [19]. While interruption of sedative therapy is an important component of assessing neurocognitive status in critically ill patients, it should also be used as a strategy to wean pharmacologic therapy when clinical condition permits. In the landmark trial conducted by Kress et al [17], sedative therapy was discontinued if patients remained comfortable following the daily interruption; if ongoing sedation was required, it was restarted at half the previous dose and only up-titrated based on patient need.

References

Monday, June 4, 2012

Innovation or impersonation: low-dose dopamine in heart failure

Although renal dysfunction is commonly associated with the use of aggressive diuretic therapy in patients with acute decompensated heart failure (ADHF), strategies for preventing or overcoming this phenomenon remain a considerable challenge. Low-dose dopamine has been proposed as a novel approach for promoting diuresis in patients with ADHF, but its use in this setting remains controversial. Enthusiasm for this strategy was renewed when preliminary results from the DAD-HF trial were announced in early 2010; however, when full details of the trial became available later that year [1], many clinicians felt that the trial's limitations outweighed many of its proposed benefits.

In DAD-HF, 60 patients with ADHF of New York Heart Association (NYHA) Class IV severity were randomized to one of two treatment arms: (1) a high-dose furosemide infusion (20 mg/hour), or (2) a low-dose furosemide infusion (5 mg/hour) plus dopamine infusion (5 mcg/kg/min). Patients with severe renal impairment (GFR < 30 mL/min) were excluded. Prior to being allocated to one of the two treatment arms, all patients received a 40-mg intravenous bolus of furosemide. All study infusions were discontinued after a total of 8 hours and further management was left to the discretion of the attending physician.

No differences in total urine output (UOP) or dyspnea symptoms were observed at 8 hours. Significant differences in the incidence of worsening renal function (defined as > 0.3 mg/dL rise in serum creatinine at 24 hours) were observed in the high-dose furosemide group (30%, compared to 6.7% in the low-dose furosemide plus dopamine group, p = 0.042); however, a difference was not observed when a second definition for worsening renal function was used (> 20% decrease in GFR at 24 hours). When patients were observed for the duration of hospitalization, results were reversed; no differences were observed when the change in serum creatinine was used to define renal dysfunction, whereas a greater percentage of patients in the high-dose furosemide group had worsening renal function when GFR was used (69.2% versus 34.6%, p = 0.025). No differences in length of stay, readmission rate, or mortality were observed.

Although the results of DAD-HF are not exactly compelling, the authors go on to make several bold statements about the proposed renoprotective benefits of dopamine. However, the limitations of this trial do warrant further discussion, as I believe they significantly influence how the results should be interpreted and applied to clinical practice.

The trial was small (n = 60) and included a considerably homogenous patient population. The hemodynamic effects of dopamine in ADHF are a function of disease severity [2], so it is unknown how patients of other NYHA classes would respond to this management strategy;
The study protocol only ran for a total of 8 hours. Given the pharmacokinetics of intravenous furosemide, a significant portion of the clinical response at 8 hours (i.e., when urine output was recorded) can attributed to the initial intravenous bolus dose (which was administered to all patients), and perhaps the first several hours of the continuous furosemide infusion. Given the impaired clearance in patients with ADHF, it is unlikely that many patients were at steady state when the infusions were discontinued and the 8-hour assessment of urine output was performed.
Assessments of efficacy were performed after 8 hours while assessments of safety were made after 24 hours (and throughout the course of hospitalization). However, no information was provided on how patients were managed after their 8-hour protocols were completed (left to the discretion of the attending physician), which could have influenced these results.
The investigators did not include a low-dose furosemide only group (i.e., no dopamine), so it is unknown whether the differences observed in DAD-HF can be attributed to dopamine per se, or the lower dose of furosemide used in this group. As the authors note in their discussion, the results of the DOSE trial demonstrated that higher doses of furosemide (whether administered as an intermittent intravenous bolus or continuous infusion) are clearly associated with worsening renal function in patients with ADHF [3]. The authors attempt to discount the possibility that their findings was due to lower doses of furosemide alone by claiming that the addition of dopamine provided a comparable degree of diuresis as high-dose furosemide but without the subsequent worsening of renal function. However, as described above, assessments of efficacy in DAD-HF were performed at 8 hours (versus 72 hours in DOSE), so this hardly seems like a fair comparison.
The dose of dopamine (5 mcg/kg/min) used in DAD-HF is hardly "low-dose" by any definition. In previous studies, the doses of dopamine thought to confer renoprotection are < 2 mcg/kg/min, where activation of renal dopaminergic receptors predominate. However, at doses of 3-5 mcg/kg/min, dopamine is known to exert positive inotropic effects, where improvements in renal blood flow can also be attributed to improved cardiac output. Even in the studies cited by the authors, the doses at which dopamine improved renal hemodynamics were also those at which cardiac index was highest [2]. In other words, it cannot be said from the present study that dopamine at 5 mcg/kg/min offers any greater degree of renoprotection than if a traditional inotrope (e.g., dobutamine, milrinone) had been used.
Finally, only surrogate markers (e.g., serum creatinine, GFR calculations) were used to determine changes in renal function. While these markers are certainly the most practical to measure and have been correlated with clinical endpoints in other investigations, it seems difficult to justify their use in some of the bold statements made by the authors of DAD-HF. No comments are made on more clinically relevant measures of renal dysfunction (e.g., time to continuous renal replacement therapy [CRRT]), or whether changes in serum creatinine or GFR were still present at 30 or 60 days after randomization. While the authors present an excellent review on the pharmacologic effects of dopamine, if these effects do not translate into meaningful improvements in clinical outcomes, what difference dose it make?

Therefore, based on the limitations highlighted above (and no additional evidence to the contrary), I tend to recommend against using dopamine as a strategy for preventing or ameliorating the renal dysfunction associated with aggressive diuretic therapy. While the adverse effects of dopamine at doses of < 5 mcg/kg/min are arguably low, I do believe it often delays initiation of more appropriate management strategies when necessary (i.e., inotropes, CRRT) while we wait to see if it will work. That being said, I am not entirely opposed to the use of dopamine in patients with ADHF, as there are some scenarios where it has its advantages:

Patients with mixed cardiogenic/septic shock or in whom the primary contributor to shock is unknown;
Patients who require inotrope therapy but have labile or tenuous blood pressures and in whom central hemodynamic parameters are unknown (i.e., scenarios in which dobutamine or milrinone may be a riskier choice); in this situation, dopamine may serve as a temporary measure until blood pressures improve (or a pulmonary artery catheter can be placed), or as a bridge to initiating a traditional inotrope;
Patients whose hemodynamics are especially sensitive to changes in intravascular fluid volume, such as those associated with aggressive diuresis or intermittent hemodialysis.

Otherwise, I am not sure that DAD-HF contributes much to our knowledge on the management of patients with ADHF or how we might avoid renal injury in the context of aggressive diuretic therapy. Additionally, based on the reasons outlined above, I am not sure it is an appropriate strategy to attempt in patients with ADHF simply because we do not yet have a more proven alternative.

For a discussion on the myths of low-dose dopamine in critically ill patients with renal dysfunction (i.e., in the absence of ADHF), please see this post.

References

Friday, May 25, 2012

Overcoming the drug shortage challenge: a call to action

For providers working in hospital and health-system settings, drug shortages have become an unfortunate reality of every day practice. While the lay media has placed much of its emphasis on cancer chemotherapy and other drugs used in oncology practice, cardiology and critical care medicine are two areas that have been hit especially hard by drug shortages. In the last six months, the following intravenous medications have been subject to extended periods of limited to no availability:

Analgesics (fentanyl, morphine)
Antibiotics (aminoglycosides, fluoroquinolones)
Antihypertensives (hydralazine)
Antiarrhythmics (lidocaine, procainamide)
Beta blockers (labetalol, metoprolol)
Benzodiazepines (diazepam, lorazepam, midazolam)
Calcium channel blockers (diltiazem)
Loop diuretics (furosemide, bumetanide)
Nutrition and electrolytes (magnesium, phosphate, sodium bicarbonate)
Vasopressors (epinephrine, vasopressin)
... and many more (antiemetics, other sedatives)

The underlying reasons for these critical drug shortages are complex and multifactorial, and which ones deserve the most blame are largely subject to one's political and economic views. Some have argued that the shortages are a result of failures in the free market, i.e., when a manufacturer decides that a product is no longer profitable, they can cease production without regard to whether patients depend on it or not. Others argue that the regulatory environment enforced by the US Food & Drug Administration (FDA) is too strict and has made it challenging for companies to stay competitive. The truth is probably somewhere in between, but no matter how you feel about the issue, the consequences are the same: drug shortages significantly impact patient care.

On Thursday, the Senate passed the Food and Drug Administration Safety and Innovation Act of 2012 (S.3187), which reauthorizes the Prescription Drug User Fee Act (PDUFA) and includes several provisions designed to help alleviate drug shortages, including:

Expedited review of generic drug approvals;
Incentives for new generic drug manufacturers to enter the market; and,
Early notification to FDA if an interruption in production is anticipated or production is being discontinued altogether.

The House is expected to vote on their version of the bill next week. In the meantime, I encourage you to contact your representatives and urge their support for PDUFA, including those provisions specifically designed to address drug shortages. If you are unsure who these individuals are for your area, please use this free resource developed by the American Society of Health-System Pharmacists (ASHP). Also, for general information and tips for political advocacy, see this toolkit developed by the American Pharmacists Association (APhA).

Saturday, May 19, 2012

Azithromycin in patients with cardiovascular disease

In a study published in the New England Journal of Medicine earlier this week, investigators observed an increased risk of sudden cardiac death associated with the macrolide antibiotic azithromycin (Zithromax^®) [1]. Azithromycin is considered a first-line option in the management of several types of upper respiratory tract infections and is one of the most widely-prescribed antibiotics in the US. The results of the trial by Ray, et al prompted the US Food & Drug Administration (FDA) to issue a safety alert addressing the use of azithromycin in patients with cardiovascular disease and further investigation is currently underway.

Using electronic medical records and prescription-use data from patients enrolled in the Tennessee Medicaid program, investigators found a small but statistically significant increase in the risk of cardiovascular death (hazard ratio 2.88, 95% CI 1.79 - 4.63, p < 0.001) when azithromycin (5-day course) was compared to no antibiotic therapy, a result that was significant both in terms of sudden cardiac death as well as other types of cardiovascular death. When compared to amoxicillin, the risk of cardiovascular death associated with azithromycin was similarly increased (hazard ratio 2.49; 95% CI 1.38 to 4.50; p = 0.002).

While there are some limitations with the use of any retrospective analysis, the findings do call into question the widely-held notion -- one that I believed until now -- that azithromycin is less cardiotoxic than other macrolides (clarithyomcin, erythromycin), where the risks of sudden cardiac death are fairly well-established. Given the known association of these agents with QT prolongation and the types of deaths observed (i.e., sudden cardiac death), the most likely etiology is a disturbance in cardiac conduction that results in a fatal ventricular arrhythmia. Similar risks have been attributed to the respiratory fluoroquinolones levofloxacin and moxifloxacin, which further limits the antibiotic choices available in this patient population.

So, in light of this new evidence, how does one manage the risk of cardiovascular death associated with azithromycin? It would be unreasonable to avoid azithromycin in all patients with cardiovascular disease; however, the results of the present study (not to mention the litigious nature of the current health care environment) should at least warrant a more careful consideration of the risks and benefits of azithromycin use.

Who is likely at risk?
The increased risk of sudden cardiac death observed in the present study likely does not apply to every patient with cardiovascular disease, especially those with milder forms (e.g., hypertension) or those who only have risk factors (e.g., dyslipidemia) for more advanced forms of cardiovascular disease. Those at greatest risk likely include:

Patients with a recent myocardial infarction, especially those with new-onset heart failure and those not receiving beta blockers (which reduce the risk of ventricular arrhythmias and sudden cardiac death in this population)
Patients with advanced heart failure, especially those who have not yet received an automatic implantable cardioverter-defibrillator (AICD)
Patients with severe electrolyte abnormalities (e.g., hypokalemia, hypomagnesemia), which are often associated with the use of chronic high-dose diuretic therapy
Patients taking anti-arrhythmic medications with known risk of QT prolongation and torsade de pointes (e.g., dofetilide, flecainide); for a list of drugs associated with QT prolongation (by risk category), please see this resource developed by the Arizona Center for Education on Research and Therapeutics

Alternative strategies in high-risk patients
If the risk of sudden cardiac death is thought to outweigh the benefit of azithromycin therapy in an individual patient, alternative antibiotics should be considered. By far, the most common indication for azithromycin is in the management of community-acquired pneumonia (CAP), where it is recommended as monotherapy in patients with no risk factors for drug-resistant Streptococcus pneumoniae, or in combination with a beta lactam in patients with comorbidities and/or risk factors for drug-resistant pathogens (e.g., chronic disease, immunosuppression, recent antibiotic use, etc) [2]. Some alternatives to consider include:

In the lowest risk population (i.e., minimal structural heart disease and no additional co-morbidities) for whom azithromycin monotherapy would have been considered, doxycycline alone is a reasonable alternative
For moderate risk patients, a combination beta lactam / beta lactamase inhibitor (e.g., amoxicillin/clavulanic acid) or second to third generation cephalosporin (cefuroxime, cefpodoxime) is likely adequate; for higher risk patients (i.e., in whom the addition of a macrolide would have been considered), addition of doxycycline is a reasonable alternative
Azithromycin is often added as adjunct therapy in patients who are at risk for atypical pathogens (e.g., Chlamydia pneumoniae, Mycoplasma pneumoniae, Legionella species), which include those with chronic pulmonary disease, long-term immunosuppression, residence in long-term care facilities, etc; in these patients, the addition of doxycycline is a reasonable alternative
Alternatives are more limited in patients with penicillin allergies; if the allergy to penicillin is reported as "rash", "upset stomach" or similar mild reactions (i.e., not true Type I hypersensitivity reactions), use of a second or third generation cephalosporin (+/- doxycycline) is reasonable, as the reported cross-reactivity with penicillins is around 5-10% or less. For true beta lactam allergies, options are further limited, as the respiratory fluoroquinolones (usually the first-line alternative in patients with penicillin allergies) have cardiotoxicities that are comparable to the macrolides; in these patients, monotherapy with doxycycline may be effective, but more thoughtful consideration as to risks and benefits of azithromycin or fluoroquinolone therapy is probably warranted, including whether additional monitoring (e.g., ambulatory ECG) should be performed
When used as part of the management of CAP, some advocate the use of loading/higher doses of doxycycline (e.g., 200 mg twice daily for at least the first day) in order to achieve adequate serum concentrations early in the treatment course [3]; given the low toxicity profile of short doxycycline courses, this strategy is probably reasonable

While the increased risk of sudden cardiac death observed with azithromycin significantly limits the choice of antibiotics used for the management of CAP in the ambulatory care setting, reasonable alternatives exist for the vast majority of cases. Also on the bright side, these findings should lead to a more careful evaluation of the risks and benefits of azithromycin therapy in this high-risk patient population, which many would argue is a step in the right direction anyway.

Acknowledgment: Thanks goes to Emily Heil, PharmD, BCPS, a clinical pharmacy specialist in infectious diseases at the University of Maryland Medical Center, who reviewed and made suggestions to the above recommendations.

References

Sunday, May 13, 2012

Rivaroxaban management guideline posted on ClotConnect

A guideline we developed for the practical management of rivaroxaban at our institution (click here for the PDF) was recently posted on ClotConnect, an educational website for patients and health care providers developed by the Hemophilia and Thrombosis Center at the University of North Carolina at Chapel Hill. Rivaroxaban is an oral factor Xa inhibitor indicated for the prevention of stroke and systemic embolism in patients with non-valvular atrial fibrillation as well as for the prevention of venous thromboembolism following orthopedic surgery. In the guideline, we provide recommendations on the following:

Considerations for initiation of therapy (dosing and administration, monitoring parameters, cost considerations)
Conversions to and from other anticoagulants, including unfractionated heparin, low molecular weight heparin, fondaparinux, dabigatran, and warfarin
Perioperative management
Management of hemorrhagic events
Patient education

I should disclose that many of the recommendations included in this document have been extrapolated from pharmacokinetic and pre-clinical investigations, as there is little other evidence in the literature to help guide clinicians in the practical management of rivaroxaban. As such, the content is subject to evolve significantly as we gain more experience with it and the other new oral anticoagulants. In the meantime, we hope this resource will be useful in assisting other institutions and individual providers who are already using rivaroxaban in clinical practice.

Special thanks goes to all the other collaborators on this project, including Melissa Hunter and Stacy Miller, who were final year student pharmacists at the time of its creation; Abbie Miller, PharmD, BCPS; and Stephan Moll, MD.

Saturday, May 12, 2012

Chlorthalidone versus hydrochlorothiazide: maybe it matters after all

At a journal club meeting earlier this week, we discussed a recently published meta-analysis/systemic review comparing chlorthalidone and hydrochlorothiazide (HCTZ) for the reduction of major cardiovascular events in patients with hypertension [1]. Controversy surrounding the widespread use of HCTZ over chlorthalidone seems to be a favorite theme among advocates for evidence-based medicine -- after all, chlorthalidone (rather than HCTZ) was the agent featured in the landmark ALLHAT trial of patients with hypertension, where improvements in cardiovascular endpoints were observed when it was compared to amlodipine and lisinopril -- all at a fraction of the cost [2]. Although the efficacy of chlorthalidone has been extrapolated to HCTZ, use of the latter has become far more widespread, where prescriptions for HCTZ outnumber those for chlorthalidone by as much as 20-fold [1]. This is likely also the result of the minimal cost of HCTZ -- currently included on the Wal-Mart $4 list and other discount programs -- and its inclusion as a combination product with nearly every other antihypertensive on the market. However, as many have suspected for a long time, there may be merit to the use of chlorthalidone over HCTZ, despite the lower cost and convenience of HCTZ.

Because no trial has directly compared chlorthalidone and HCTZ, Roush, et al conducted a retrospective review of nine randomized trials (n > 50,000) as well as drug-adjusted and blood pressure-adjusted meta-analyses comparing these two agents in the management of hypertension. Compared to HCTZ, chlorthalidone reduced the risk of cardiovascular events by 21% (95% CI 12-28%, p = 0.0001) and risk of heart failure by 23% (95% CI 2-39%, p = 0.032) -- results that were independent of actual blood pressure control. Impressively, the number-needed-to-treat to prevent a single cardiovascular event with chlorthalidone (compared to HCTZ) over a 5-year period was only 27. The investigators attribute these differences to several possible factors, including effects of chlorthalidone unrelated to its vasodilatory properties, as well as known pharmacokinetic differences between chlorthalidone and HCTZ.

Although there are usually other compelling indications that warrant the use of alternative anti-hypertensives in my patient population (e.g., ACE inhibitors in patients with established coronary artery disease, heart failure, diabetes, or chronic kidney disease), there are some instances where the decision to initiate either chlorthalidone or HCTZ arises (e.g., hypertensive urgency, uncontrolled hypertension on maximum doses of other therapies). In these scenarios, when cost or pill burden is not an issue, I recommend chlorthalidone over HCTZ, not so much for the academic purity of it but because of its pharmacokinetic advantages. Chlorthalidone has a terminal half-life of over 40 hours -- one of the longest of any anti-hypertensive currently available. Because patients with hypertension rarely "feel" the symptoms of their disease, daily compliance with anti-hypertensive medications can be a significant issue. With such a long terminal half-life, even if a patient remembers to take chlorthalidone every other day, they are likely to derive some anti-hypertensive benefit.

The only drawback with this strategy is cost and pill burden. Although chlorthalidone is generic and relatively inexpensive, it is not available on any of the retail discount programs (of which I am aware). The only exception is in a combination product with atenolol, but I usually have reasons for using an alternative beta blocker (e.g., metoprolol, carvedilol, etc), so this is not usually an option. Moreover, compliance decreases as the number of medications increases, so the limited number of combination products available makes chlorthalidone an entirely separate prescription to fill and another pill to take.

However, if neither of these issues is the case, it seems that the analysis conducted by Roush, et al, only adds further evidence to support the use of chlorthalidone over HCTZ.

Wednesday, March 28, 2012

New monoclonal antibody REGN727: more effective than statins?

Speaking of surrogate endpoints, I ran across the following headline in Daily Briefing, an e-newsletter I receive as part of my membership in the American Society of Health-System Pharmacists (ASHP): Experimental Drug May Be More Effective Than Statins.

The headline links to several published reports (including this one on theheart.org) about REGN727, an investigational human monoclonal antibody to proprotein convertase subtilisin/kexin 9 (PCSK9). According to the article, PCSK9 is a serum protease involved in the breakdown of hepatic low-density lipoprotein (LDL) receptors. As a result of antibody-mediated inhibition of PCSK9 by REGN727, hepatic LDL receptors are spared, leading to subsequent reductions in serum LDL concentrations. At the 2012 American College of Cardiology meeting earlier this week, lead investigators said REGN727 could be a potential "game changer" in the management of dyslipidemia, although a significant amount of research would still be required before the agent becomes available in the US.

In these early phase I trials, a subcutaneous injection of REGN727 was effective at reducing LDL concentrations by up to three-fourths, even in patients already taking maximally-tolerated doses of atorvastatin. Notably, the studies were conducted in healthy volunteers and were not powered to detect differences in clinical outcomes (e.g., cardiovascular morbidity and mortality).

While the drug may be effective at reducing serum LDL concentrations, questions remain as to whether these effects impart significant improvements in cardiovascular outcomes, as we have seen several recent examples where this correlation does not exist. One investigation in particular is ARBITER 6-HALTS, which compared extended-release niacin (Niaspan^®) to ezetimibe (Zetia^®) and found that, despite being more effective at reducing LDL concentrations, ezetimibe was inferior to niacin at reducing cardiovascular events [1].

So, is this new agent promising? Yes. Is it worth further investigation? Of course. Alternatives are limited for patients who continue to have progressive atherosclerosis despite maximally-tolerated doses of statins and other antidyslipidemic medications. But to say REGN727 "may be more effective" than statins is a bit of a stretch at this time. Statins do reduce serum LDL concentrations, but they also improve cardiovascular outcomes independently of their effects on LDL, even in patients with cholesterol in the accepted "normal" range. Whether REGN727 is also able to do this -- independently of its effects on serum LDL concentrations -- is yet to be seen, but is what I think will determine its overall clinical utility in the management of atherosclerotic disease.

References

Taylor AJ, Villines TC, Turco M, et al. Extended-release niacin or ezetimibe and carotid intima-media thickness. N Engl J Med. 2009 Nov 26;361(22):2113-22.

Tuesday, March 27, 2012

Surrogate markers: reflections of efficacy or only graven images?

I ran across this article the other day, in which Richard Lehman, a British physician and blogger for BMJ, reviews highlights from several popular medical journals. At the end, he cites an article entitled The Idolatry of the Surrogate [2], where Yudkin, et al present an excellent discussion of the controversies surrounding the use of surrogate markers of efficacy versus hard clinical outcomes. The article by Yudkin, et al uses diabetes as an example, but the same argument could be made for many of the markers commonly used in cardiovascular clinical practice (e.g., low-density lipoprotein, carotid intima media thickness, etc).

One of the best parts of Lehman's piece is his citation of Yudkin's Ten Commandments, a section that did not actually make it into the Idolatry piece due to concern over potential religious sensitivities. I found its omission unfortunate, as it is one of the best anecdotes on clinical practice I have read in a while and one I wanted to share:

The New Therapeutics: Ten Commandments

Thou shalt treat according to level of risk rather than level of risk factor.
Thou shalt exercise caution when adding drugs to existing polypharmacy.
Thou shalt consider benefits of drugs as proven only by hard endpoint studies.
Thou shalt not bow down to surrogate endpoints, for these are but graven images.
Thou shalt not worship Treatment Targets, for these are but the creations of Committees.
Thou shalt apply a pinch of salt to Relative Risk Reductions, regardless of P values, for the population of their provenance may bear little relationship to thy daily clientele.
Thou shalt honour the Numbers Needed to Treat, for therein rest the clues to patient-relevant information and to treatment costs.
Thou shalt not see detailmen, nor covet an Educational Symposium in a luxury setting.
Thou shalt share decisions on treatment options with the patient in the light of estimates of the individual’s likely risks and benefits.
Honour the elderly patient, for although this is where the greatest levels of risk reside, so do the greatest hazards of many treatments.

Those of you who follow my blog regularly know that this topic is one of my favorite to debate, as I think it is one of the most controversial in all of professional practice. In the vast majority cases, I tend to side with Yudkin's arguments -- I believe the above considerations lead one to think more critically about a patient's overall risk and the potential for benefit (or additional risk) with any given intervention. Unfortunately, this is often met with some initial resistance, as these arguments rarely fall in line with recognized practice guidelines.

But, as the analysis by Smith, et al in 2009 pointed out, only about a tenth of the recommendations made in cardiology guidelines are based on randomized controlled trials -- the remaining majority are based on small or limited patient populations and over half from expert consensus or standards of care [2].

... which only makes this topic worth debating again and again.

References

Tuesday, March 20, 2012

Optimizing strategies for preventing contrast-induced nephropathy

Unfortunately, the motivation for writing this entry was not for the purpose of sharing the findings of a new study, but rather the critical nationwide shortage of sodium bicarbonate, which has left us carefully scrutinizing our strategies for preventing contrast-induced nephropathy (CIN) in patients undergoing cardiac catheterization. For those of you working in a health-system setting, your practice has probably been severely hampered by the growing number of shortages across several therapeutic areas -- hopefully the results from the following study will alleviate some concerns about the need for sodium bicarbonate in the prevention of CIN.

Contrast-induced nephropathy is an unfortunate risk of coronary angiography and it is especially high among patients with existing renal impairment. Given the extensive use of angiography to evaluate coronary artery disease, strategies to prevent or reduce the risk of CIN have been the subject of ongoing research and unfortunately, the options remain limited. Oral N-acetylcysteine is no longer recommended (see a summary of the 2012 guidelines for percutaneous coronary intervention here), which leaves intravenous fluid hydration as one of the few remaining alternatives for renoprotection in the setting of cardiac catheterization. However, the optimal agent recommended for this purpose (i.e., normal saline versus sodium bicarbonate) has not been well-established.

In a study published in late January, Klima, et al [1] randomized 258 patients with renal impairment undergoing intravascular contrast procedures to one of three hydration strategies:

Sodium chloride 0.9% (i.e., normal saline) 1 mL/kg/h for at least 12 hours before and after the procedure
Sodium bicarbonate (166 mEq/L) 3 mL/kg for 1 hour before and 1 mL/kg/h for 6 hours after the procedure; or,
Sodium bicarbonate (166 mEq/L) 3 mL/kg over 20 minutes before the procedure plus oral sodium bicarbonate (500 mg per 10 kg)

Sodium chloride was shown to be superior to sodium bicarbonate at reducing the primary endpoint of change in estimated glomerular filtration rate (eGFR); it was also more effective at reducing the development of CIN, defined as an increase in serum creatinine (SCr) of > 25% or an increase of > 0.5 mg/dL from baseline. Moreover, the shorter infusion of sodium bicarbonate (20 minutes) plus oral sodium bicarbonate was just as effective as the 7-hour infusion for improving these same endpoints. No differences were observed in long-term morbidity and mortality or progression to renal replacement therapy.

But does the present study indicate superiority of sodium chloride over sodium bicarbonate?

Not really.

I think the results of this trial only confirms what we've suspected for a long time -- the more hydration around a procedure, the lower the risk of CIN. That being said, I do think this adds some helpful information to the longstanding debate over which strategy is best for reducing the risk of CIN in the setting of intravascular contrast procedures.

Previous investigations have been rife with limitations. Furthermore, differences in the types of procedures performed, study populations, and types of contrast dyes utilized have only further complicated efforts to identify a conclusive strategy for renoprotection. In a recent meta-analysis by Meier, et al, sodium bicarbonate was found to be more effective at reducing CIN; however, effects were less dramatic among patients receiving elective procedures or those receiving iso-osmolar contrast dyes [2]. Another commonly cited study was conducted by Merten et al, where the 7-hour sodium bicarbonate infusion used in the present study was shown to be superior to a 7-hour infusion of sodium chloride [3]. In both instances, no differences in mortality or progression to renal replacement therapy were observed.

I find some comfort in the results of this trial, as it provides us with an alternative for preventing CIN in the setting of a critical nationwide sodium bicarbonate shortage. However, one remaining question is how long is long enough? Twelve hours of sodium chloride before and after an intravascular contrast procedure may be more effective than 7 hours of sodium bicarbonate (1 hour before and 6 hours after), but what if the need for catheterization was more urgent (i.e., 4-6 hours or less prior to procedure)? For emergent procedures (i.e., where only < 1 hour of renoprotective measures can be provided prior to procedure), I think we are still compelled to use sodium bicarbonate in the absence of any data stating otherwise.

Secondly, what about patients with moderate to severe heart failure, where excessive intravenous fluid administration could result in clinical decompensation? Patients with New York Heart Association Class III-IV heart failure were excluded from the present study, and rightly so -- a patient weighing only 70 kg would have received > 1.5 L within a 24-hour period, i.e., definitely enough fluid to promote an exacerbation in a patient with existing heart failure at baseline. In this particular scenario, I think sodium bicarbonate would still be the most attractive option for renoprotection.

Finally, another interesting finding from the present study is that the shorter course of sodium bicarbonate (20 minutes prior to procedure) combined with oral sodium bicarbonate was just as effective as the 7-hour infusion. I imagine this strategy would be an attractive option in a number of scenarios, including elective and outpatient intravascular contrast procedures.

At the very least, I think the study by Klima, et al provides us with evidence to support the use of sodium chloride in patients for whom an intravascular contrast procedure is not urgent or emergent. For these latter scenarios, sodium bicarbonate still appears to be the better option, although it may not be one we can actually use as a result of the ongoing shortage nationwide. In this case, it appears our only alternative is to hydrate with sodium chloride as much as possible prior to procedure and at least 12 hours afterwards and then hope for the best.

For more information on drug shortages and strategies for addressing them, please see the Drug Shortages Resource Center created by the American Society for Health-System Pharmacists.

References