Accumulating evidence suggests that perioperative treatment with 3‐hydroxy‐3‐methylglutaryl coenzyme‐A (HMG‐CoA) reductase inhibitors (or, statins) reduces the incidence of cardiovascular events during noncardiac surgery.16 This evidence has lead the European Society of Cardiology (ESC) and American College of Cardiology Foundation/American Heart Association (ACCF/AHA) to endorse the use of perioperative statins in patients already on this treatment or those at high‐risk of cardiovascular events.7, 8

However, statins are available only in oral formulation. Consequently, prolonged bowel recovery or clinical instability may interfere with use during surgery. Furthermore, many clinicians may not recognize the imperative of postoperative statin resumption, viewing them principally as lipid‐lowering entities and not as agents of perioperative benefit. Failure to resume statins postoperatively can be catastrophic, as the ensuing inflammation and thrombosis frequently culminates in myocardial infarction (MI) or death.9, 10

In this article, we review the potent anti‐inflammatory properties of statins and their role in preventing perioperative cardiac events. We outline the biochemical basis for perioperative statin benefit, summarizing the basic, clinical, and experimental evidence regarding statin withdrawal. We conclude by presenting strategies to avert postoperative statin cessation and outline a research agenda dedicated to informing this practice.

METHODS

We performed a literature search using MEDLINE via Ovid (1946present), EMBASE (1946present), Biosis (1926present), and Cochrane CENTRAL (1960present). We used Boolean logic to search for key terms including statins, 3‐hydroxy‐3‐methylglutaryl CoA reductase inhibitors, death, MI, stroke, acute coronary syndrome (ACS), and statin withdrawal or cessation. All studies published in full‐text or abstract form were included. A total of 489 articles were retrieved by this search (last updated March 15, 2012). For this narrative review, we focused on studies that examined adverse outcomes associated with statin withdrawal.

BIOCHEMICAL BASIS OF STATIN PLEIOTROPICITY

The nonlipid‐lowering or pleiotropic properties of statins are especially valuable in the perioperative setting.16, 11 Perioperative cardiac complications occur due to oxygen supply:demand mismatch, vascular inflammation, or a combination of these states. A significant perisurgical catecholamine surge produces unopposed sympathetic effects,12 increasing the risk of rupture of vulnerable coronary plaques, thrombus formation, and adverse cardiac events.13, 14 Similarly, augmented inflammatory responses and increased circulating coagulation factors further predispose to a hazardous perioperative milieu.15 Statins attenuate this vascular inflammatory response by suppressing the synthesis of mevalonate by inhibiting HMG‐CoA reductase. Suppression of mevalonate synthesis reduces the bioavailability of 2 important isoprenoid molecules: farnesyl‐pyrophosphate and geranylgeranyl‐pyrophosphate.16 Diminution of these isoprenoid intermediaries leads to reductions in the active intracellular signaling molecules Ras, Rho, and Rac, which play critical roles in vascular reactivity, endothelial function, and coagulation and inflammatory pathways.1723 The cumulative effect of these cellular changes is diminished inflammation during periods of surgical stress (Figure 1).

Figure 1

While the perioperative pleiotropicity of statins is of inherent clinical value, several studies have shown that these effects are lost and even reversed when statins are withdrawn.2428 During statin treatment, absence of isoprenoid intermediaries induces cytosolic accumulation of nonactivated Rho and Rac proteins. Abrupt cessation of statins activates Rho/Rac‐kinase pathways, leading to unregulated inflammation, platelet hyper‐activation, and endothelial dysfunction.24, 25, 28, 29 For instance, statin withdrawal in mice‐models leads to an overshoot activation of Rho, resulting in down‐regulation of endothelial nitric oxide production,25 activation of nicotinamide adenine dinucleotide phosphate (NAD[P]H)‐oxidase, and increased superoxide production.29 In another mouse‐model, statin withdrawal was associated with up‐regulation of key pro‐thrombotic molecules including platelet factor 4 and beta‐thromboglobulin.24 In human studies, a platelet hyper‐activation state (manifested by increased platelet P‐selectin expression and enhanced platelet aggregation) occurs after statin discontinuation.27 Furthermore, withdrawal of statins in patients with hyperlipidemia increases inflammatory markers such as C‐reactive protein and interleukin‐6.26 In the perioperative context, absence of these important anti‐inflammatory properties increases the risk of cardiac events.9, 10

EVIDENCE SUGGESTING BENEFIT FROM PERIOPERATIVE STATIN TREATMENT

Retrospective studies first suggested clinical benefit from perioperative statin treatment. In a case‐control study involving 2816 patients undergoing vascular surgery at Erasmus Medical Center, statin use was associated with substantially decreased postoperative mortality (adjusted odd ratio [OR] 0.22, 95% confidence interval [CI] 0.100.47).5 In a subsequent retrospective cohort study of 780,591 patients who underwent major noncardiac surgery, the risk of postoperative mortality was considerably lower among statin users (unadjusted OR 0.68, 95% CI 0.640.72) compared to patients who did not receive, or received delayed treatment with statins.3 A third retrospective study of 1163 vascular surgery patients found that statins prevented perioperative cardiac complications including death, MI, congestive heart failure, and ventricular tachyarrhythmias (OR 0.52, 95% CI 0.350.76).4

The benefit from statin treatment found in retrospective studies prompted the first double‐blinded, randomized controlled trial (RCT) of perioperative statin use. In 2004, Durazzo and colleagues1 randomized 100 statin‐naive patients scheduled to undergo elective aortic, femoro‐popliteal, or carotid surgery to receive either 20 mg of atorvastatin or placebo for 45 days. Vascular surgery was performed, on average, 31 days after randomization. Atorvastatin therapy reduced the incidence of death from cardiac causes, nonfatal acute MI, ischemic stroke, and unstable angina (26% in the placebo group vs 8% in the atorvastatin group; P = 0.031).1 Although the small size of the trial rendered it underpowered to show a mortality benefit, this remains the first RCT to demonstrate a protective perioperative effect of statins.

In the 2009 Dutch Echocardiographic Cardiac Risk Evaluation Applying Stress Echocardiography (DECREASE)‐III trial, Schouten and colleagues6 randomized 497 high‐risk, statin‐naive patients undergoing vascular surgery to receive, in addition to a beta‐blocker, either fluvastatin or placebo before surgery (median of 37 days). Postoperative myocardial ischemia (hazard ratio [HR] 0.55, 95% CI 0.340.88), and combined death from cardiovascular causes or nonfatal MI (HR 0.47, 95% CI 0.240.94), occurred less frequently in the treatment group.6 In 2009, the same group published DECREASE‐IV, a multicenter, prospective, open‐label, 2 2 factorial design trial of 1066 intermediate‐risk patients, scheduled to undergo elective, noncardiac surgery. Patients were assigned to bisoprolol, fluvastatin, combination treatment, or control therapy before surgery (median of 34 days). Although those randomized to fluvastatin demonstrated lower incidence of 30‐day cardiac death and MI than control (HR 0.65, 95% CI 0.351.10), these outcomes failed to reach statistical significance as the trial was principally powered to examine the effects of perioperative beta‐blockade.30

Using this pool of data, a meta‐analysis of 15 studies (223,010 patients) found a substantial 38% reduction in the risk of mortality after cardiac surgery (1.9% vs 3.1%; P = 0.0001) and an even greater 59% reduction in the risk of mortality following vascular surgery (1.7% vs 6.1%; P = 0.0001) with perioperative statin therapy. When including noncardiac surgery, a 44% reduction in mortality was observed (2.2% vs 3.2%; P < 0.01).2 We performed a similar meta‐analysis of 15 RCTs involving 2292 patients to determine whether perioperative statin treatment in statin‐naive patients, undergoing either cardiac or noncardiac surgery, improved clinical outcomes. Our analysis also found statistically significant reductions in the risk of MI associated with perioperative statin use in both cardiac and noncardiac surgery (risk reduction [RR] 0.53, 95% CI 0.380.74) and atrial fibrillation in statin‐naive patients undergoing cardiac surgery (RR 0.56, 95% CI 0.450.69).31 Taken together, a large volume of evidence supports the use of statins in surgical settings.

In view of this evidence, the ACCF/AHA perioperative guidelines for noncardiac surgery endorsed statins as an important risk‐reducing intervention in those undergoing noncardiac surgery, and recommended continued use in patients on chronic statin treatment scheduled for noncardiac surgery (Level of Evidence B, Class I; Benefits >>> Risk). Initiating statins in patients undergoing vascular surgery, with or without risk factors, was considered reasonable (Level of Evidence B, Class IIa; Benefits >> Risk).7 Current ESC perioperative guidelines in noncardiac surgery offer similar recommendations to those of ACCF/AHA, but differ by categorizing the recommendation to initiate statins in patients at high cardiovascular risk as a Class I recommendation.8

CLINICAL CONSEQUENCES OF STATIN WITHDRAWAL

Although statins provide important cardiac benefits, an important limitation to their perioperative use remains their oral‐only formulation. Thus, patients who are unable to resume oral intake may fail to resume treatment. Perioperative statin cessation has been hypothesized to lead to a statin withdrawal phenomenon. The evidence that supports the existence of this phenomenon comes from 3 distinct populations: ACS, ischemic stroke, and perioperative patients (Table 1).

CLINICAL INSIGHTS INTO FAILURE OF POSTOPERATIVE STATIN RESUMPTION

We hypothesize that failure to resume perioperative statins may occur for 4 cardinal reasons. First, resumption of an oral agent frequently proves clinically challenging when complications such as postoperative ileus, nausea, and vomiting peak. To date, no intravenous statin formulations are available, although phase‐I studies are currently underway.38 Second, it is not inconceivable that perioperative clinical instability may overshadow the resumption of statin treatment. Third, clinicians may also remain concerned regarding adverse effects of statins, a thought compounded by US Food and Drug Administration statin package inserts that specifically advocate for statins to be withheld during surgery. However, although the occurrence of elevated liver function tests and myopathy are theoretically important, the overwhelming majority of perioperative statin studies in noncardiac surgery have not found this to be a major occurrence.39 Nonetheless, a lack of uniform definitions and appropriate surveillance for adverse events are important limitations to this finding. In our recent systematic review, we were unable to provide refined estimates of these important side effects owing to differences in definition, variations in screening, and absence of standardized cutoffs used in studies.31 Finally, an important reason for failing to resume postoperative statins is that many physicians simply fail to recognize the perioperative importance of these agents.

STRATEGIES TO IMPROVE PERIOPERATIVE STATIN RESUMPTION

Using the existing evidence, we propose the following 4 clinical strategies to assist in avoiding a statin withdrawal state.

Preoperative Transition to Extended Release Statin Formulations

An innovative approach to minimizing statin withdrawal involves preoperative transition to an extended‐release statin formulation. This strategy may be of particular value in patients where prolonged bowel nonavailability is likely, such as those undergoing gastrointestinal surgery, or when prolonged postoperative dietary restriction (eg, NPO [nil per os]: nothing by mouth) status is expected (Figure 2).

Figure 2

CONCLUSIONS AND FUTURE DIRECTIONS

Sudden withdrawal of perioperative statins results in adverse clinical outcomes. Individuals engaged in the care of patients during surgery such as hospitalists, anesthesiologists, and surgeons must become more cognizant of a statin withdrawal state.

An important limitation associated with the study of perioperative statin withdrawal remains the ambiguity regarding the extent of the problem in the United States. Therefore, a logical first step could be the use of infrastructure within the National Surgical Quality Improvement Program (NSQIP) to understand the epidemiology of perioperative statin use and consequences associated with statin discontinuation.44 Mandating such quality reporting could easily be built into current NSQIP performance metrics. These data would help inform a research agenda targeting patients that experience statin withdrawal and strategies most likely to prevent it.

Accumulating evidence suggests that perioperative treatment with 3‐hydroxy‐3‐methylglutaryl coenzyme‐A (HMG‐CoA) reductase inhibitors (or, statins) reduces the incidence of cardiovascular events during noncardiac surgery.16 This evidence has lead the European Society of Cardiology (ESC) and American College of Cardiology Foundation/American Heart Association (ACCF/AHA) to endorse the use of perioperative statins in patients already on this treatment or those at high‐risk of cardiovascular events.7, 8

However, statins are available only in oral formulation. Consequently, prolonged bowel recovery or clinical instability may interfere with use during surgery. Furthermore, many clinicians may not recognize the imperative of postoperative statin resumption, viewing them principally as lipid‐lowering entities and not as agents of perioperative benefit. Failure to resume statins postoperatively can be catastrophic, as the ensuing inflammation and thrombosis frequently culminates in myocardial infarction (MI) or death.9, 10

In this article, we review the potent anti‐inflammatory properties of statins and their role in preventing perioperative cardiac events. We outline the biochemical basis for perioperative statin benefit, summarizing the basic, clinical, and experimental evidence regarding statin withdrawal. We conclude by presenting strategies to avert postoperative statin cessation and outline a research agenda dedicated to informing this practice.

METHODS

We performed a literature search using MEDLINE via Ovid (1946present), EMBASE (1946present), Biosis (1926present), and Cochrane CENTRAL (1960present). We used Boolean logic to search for key terms including statins, 3‐hydroxy‐3‐methylglutaryl CoA reductase inhibitors, death, MI, stroke, acute coronary syndrome (ACS), and statin withdrawal or cessation. All studies published in full‐text or abstract form were included. A total of 489 articles were retrieved by this search (last updated March 15, 2012). For this narrative review, we focused on studies that examined adverse outcomes associated with statin withdrawal.

BIOCHEMICAL BASIS OF STATIN PLEIOTROPICITY

The nonlipid‐lowering or pleiotropic properties of statins are especially valuable in the perioperative setting.16, 11 Perioperative cardiac complications occur due to oxygen supply:demand mismatch, vascular inflammation, or a combination of these states. A significant perisurgical catecholamine surge produces unopposed sympathetic effects,12 increasing the risk of rupture of vulnerable coronary plaques, thrombus formation, and adverse cardiac events.13, 14 Similarly, augmented inflammatory responses and increased circulating coagulation factors further predispose to a hazardous perioperative milieu.15 Statins attenuate this vascular inflammatory response by suppressing the synthesis of mevalonate by inhibiting HMG‐CoA reductase. Suppression of mevalonate synthesis reduces the bioavailability of 2 important isoprenoid molecules: farnesyl‐pyrophosphate and geranylgeranyl‐pyrophosphate.16 Diminution of these isoprenoid intermediaries leads to reductions in the active intracellular signaling molecules Ras, Rho, and Rac, which play critical roles in vascular reactivity, endothelial function, and coagulation and inflammatory pathways.1723 The cumulative effect of these cellular changes is diminished inflammation during periods of surgical stress (Figure 1).

Figure 1

While the perioperative pleiotropicity of statins is of inherent clinical value, several studies have shown that these effects are lost and even reversed when statins are withdrawn.2428 During statin treatment, absence of isoprenoid intermediaries induces cytosolic accumulation of nonactivated Rho and Rac proteins. Abrupt cessation of statins activates Rho/Rac‐kinase pathways, leading to unregulated inflammation, platelet hyper‐activation, and endothelial dysfunction.24, 25, 28, 29 For instance, statin withdrawal in mice‐models leads to an overshoot activation of Rho, resulting in down‐regulation of endothelial nitric oxide production,25 activation of nicotinamide adenine dinucleotide phosphate (NAD[P]H)‐oxidase, and increased superoxide production.29 In another mouse‐model, statin withdrawal was associated with up‐regulation of key pro‐thrombotic molecules including platelet factor 4 and beta‐thromboglobulin.24 In human studies, a platelet hyper‐activation state (manifested by increased platelet P‐selectin expression and enhanced platelet aggregation) occurs after statin discontinuation.27 Furthermore, withdrawal of statins in patients with hyperlipidemia increases inflammatory markers such as C‐reactive protein and interleukin‐6.26 In the perioperative context, absence of these important anti‐inflammatory properties increases the risk of cardiac events.9, 10

EVIDENCE SUGGESTING BENEFIT FROM PERIOPERATIVE STATIN TREATMENT

Retrospective studies first suggested clinical benefit from perioperative statin treatment. In a case‐control study involving 2816 patients undergoing vascular surgery at Erasmus Medical Center, statin use was associated with substantially decreased postoperative mortality (adjusted odd ratio [OR] 0.22, 95% confidence interval [CI] 0.100.47).5 In a subsequent retrospective cohort study of 780,591 patients who underwent major noncardiac surgery, the risk of postoperative mortality was considerably lower among statin users (unadjusted OR 0.68, 95% CI 0.640.72) compared to patients who did not receive, or received delayed treatment with statins.3 A third retrospective study of 1163 vascular surgery patients found that statins prevented perioperative cardiac complications including death, MI, congestive heart failure, and ventricular tachyarrhythmias (OR 0.52, 95% CI 0.350.76).4

The benefit from statin treatment found in retrospective studies prompted the first double‐blinded, randomized controlled trial (RCT) of perioperative statin use. In 2004, Durazzo and colleagues1 randomized 100 statin‐naive patients scheduled to undergo elective aortic, femoro‐popliteal, or carotid surgery to receive either 20 mg of atorvastatin or placebo for 45 days. Vascular surgery was performed, on average, 31 days after randomization. Atorvastatin therapy reduced the incidence of death from cardiac causes, nonfatal acute MI, ischemic stroke, and unstable angina (26% in the placebo group vs 8% in the atorvastatin group; P = 0.031).1 Although the small size of the trial rendered it underpowered to show a mortality benefit, this remains the first RCT to demonstrate a protective perioperative effect of statins.

In the 2009 Dutch Echocardiographic Cardiac Risk Evaluation Applying Stress Echocardiography (DECREASE)‐III trial, Schouten and colleagues6 randomized 497 high‐risk, statin‐naive patients undergoing vascular surgery to receive, in addition to a beta‐blocker, either fluvastatin or placebo before surgery (median of 37 days). Postoperative myocardial ischemia (hazard ratio [HR] 0.55, 95% CI 0.340.88), and combined death from cardiovascular causes or nonfatal MI (HR 0.47, 95% CI 0.240.94), occurred less frequently in the treatment group.6 In 2009, the same group published DECREASE‐IV, a multicenter, prospective, open‐label, 2 2 factorial design trial of 1066 intermediate‐risk patients, scheduled to undergo elective, noncardiac surgery. Patients were assigned to bisoprolol, fluvastatin, combination treatment, or control therapy before surgery (median of 34 days). Although those randomized to fluvastatin demonstrated lower incidence of 30‐day cardiac death and MI than control (HR 0.65, 95% CI 0.351.10), these outcomes failed to reach statistical significance as the trial was principally powered to examine the effects of perioperative beta‐blockade.30

Using this pool of data, a meta‐analysis of 15 studies (223,010 patients) found a substantial 38% reduction in the risk of mortality after cardiac surgery (1.9% vs 3.1%; P = 0.0001) and an even greater 59% reduction in the risk of mortality following vascular surgery (1.7% vs 6.1%; P = 0.0001) with perioperative statin therapy. When including noncardiac surgery, a 44% reduction in mortality was observed (2.2% vs 3.2%; P < 0.01).2 We performed a similar meta‐analysis of 15 RCTs involving 2292 patients to determine whether perioperative statin treatment in statin‐naive patients, undergoing either cardiac or noncardiac surgery, improved clinical outcomes. Our analysis also found statistically significant reductions in the risk of MI associated with perioperative statin use in both cardiac and noncardiac surgery (risk reduction [RR] 0.53, 95% CI 0.380.74) and atrial fibrillation in statin‐naive patients undergoing cardiac surgery (RR 0.56, 95% CI 0.450.69).31 Taken together, a large volume of evidence supports the use of statins in surgical settings.

In view of this evidence, the ACCF/AHA perioperative guidelines for noncardiac surgery endorsed statins as an important risk‐reducing intervention in those undergoing noncardiac surgery, and recommended continued use in patients on chronic statin treatment scheduled for noncardiac surgery (Level of Evidence B, Class I; Benefits >>> Risk). Initiating statins in patients undergoing vascular surgery, with or without risk factors, was considered reasonable (Level of Evidence B, Class IIa; Benefits >> Risk).7 Current ESC perioperative guidelines in noncardiac surgery offer similar recommendations to those of ACCF/AHA, but differ by categorizing the recommendation to initiate statins in patients at high cardiovascular risk as a Class I recommendation.8

CLINICAL CONSEQUENCES OF STATIN WITHDRAWAL

Although statins provide important cardiac benefits, an important limitation to their perioperative use remains their oral‐only formulation. Thus, patients who are unable to resume oral intake may fail to resume treatment. Perioperative statin cessation has been hypothesized to lead to a statin withdrawal phenomenon. The evidence that supports the existence of this phenomenon comes from 3 distinct populations: ACS, ischemic stroke, and perioperative patients (Table 1).

CLINICAL INSIGHTS INTO FAILURE OF POSTOPERATIVE STATIN RESUMPTION

We hypothesize that failure to resume perioperative statins may occur for 4 cardinal reasons. First, resumption of an oral agent frequently proves clinically challenging when complications such as postoperative ileus, nausea, and vomiting peak. To date, no intravenous statin formulations are available, although phase‐I studies are currently underway.38 Second, it is not inconceivable that perioperative clinical instability may overshadow the resumption of statin treatment. Third, clinicians may also remain concerned regarding adverse effects of statins, a thought compounded by US Food and Drug Administration statin package inserts that specifically advocate for statins to be withheld during surgery. However, although the occurrence of elevated liver function tests and myopathy are theoretically important, the overwhelming majority of perioperative statin studies in noncardiac surgery have not found this to be a major occurrence.39 Nonetheless, a lack of uniform definitions and appropriate surveillance for adverse events are important limitations to this finding. In our recent systematic review, we were unable to provide refined estimates of these important side effects owing to differences in definition, variations in screening, and absence of standardized cutoffs used in studies.31 Finally, an important reason for failing to resume postoperative statins is that many physicians simply fail to recognize the perioperative importance of these agents.

STRATEGIES TO IMPROVE PERIOPERATIVE STATIN RESUMPTION

Using the existing evidence, we propose the following 4 clinical strategies to assist in avoiding a statin withdrawal state.

Preoperative Transition to Extended Release Statin Formulations

An innovative approach to minimizing statin withdrawal involves preoperative transition to an extended‐release statin formulation. This strategy may be of particular value in patients where prolonged bowel nonavailability is likely, such as those undergoing gastrointestinal surgery, or when prolonged postoperative dietary restriction (eg, NPO [nil per os]: nothing by mouth) status is expected (Figure 2).

Figure 2

CONCLUSIONS AND FUTURE DIRECTIONS

Sudden withdrawal of perioperative statins results in adverse clinical outcomes. Individuals engaged in the care of patients during surgery such as hospitalists, anesthesiologists, and surgeons must become more cognizant of a statin withdrawal state.

An important limitation associated with the study of perioperative statin withdrawal remains the ambiguity regarding the extent of the problem in the United States. Therefore, a logical first step could be the use of infrastructure within the National Surgical Quality Improvement Program (NSQIP) to understand the epidemiology of perioperative statin use and consequences associated with statin discontinuation.44 Mandating such quality reporting could easily be built into current NSQIP performance metrics. These data would help inform a research agenda targeting patients that experience statin withdrawal and strategies most likely to prevent it.

Accumulating evidence suggests that perioperative treatment with 3‐hydroxy‐3‐methylglutaryl coenzyme‐A (HMG‐CoA) reductase inhibitors (or, statins) reduces the incidence of cardiovascular events during noncardiac surgery.16 This evidence has lead the European Society of Cardiology (ESC) and American College of Cardiology Foundation/American Heart Association (ACCF/AHA) to endorse the use of perioperative statins in patients already on this treatment or those at high‐risk of cardiovascular events.7, 8

However, statins are available only in oral formulation. Consequently, prolonged bowel recovery or clinical instability may interfere with use during surgery. Furthermore, many clinicians may not recognize the imperative of postoperative statin resumption, viewing them principally as lipid‐lowering entities and not as agents of perioperative benefit. Failure to resume statins postoperatively can be catastrophic, as the ensuing inflammation and thrombosis frequently culminates in myocardial infarction (MI) or death.9, 10

In this article, we review the potent anti‐inflammatory properties of statins and their role in preventing perioperative cardiac events. We outline the biochemical basis for perioperative statin benefit, summarizing the basic, clinical, and experimental evidence regarding statin withdrawal. We conclude by presenting strategies to avert postoperative statin cessation and outline a research agenda dedicated to informing this practice.

METHODS

We performed a literature search using MEDLINE via Ovid (1946present), EMBASE (1946present), Biosis (1926present), and Cochrane CENTRAL (1960present). We used Boolean logic to search for key terms including statins, 3‐hydroxy‐3‐methylglutaryl CoA reductase inhibitors, death, MI, stroke, acute coronary syndrome (ACS), and statin withdrawal or cessation. All studies published in full‐text or abstract form were included. A total of 489 articles were retrieved by this search (last updated March 15, 2012). For this narrative review, we focused on studies that examined adverse outcomes associated with statin withdrawal.

BIOCHEMICAL BASIS OF STATIN PLEIOTROPICITY

The nonlipid‐lowering or pleiotropic properties of statins are especially valuable in the perioperative setting.16, 11 Perioperative cardiac complications occur due to oxygen supply:demand mismatch, vascular inflammation, or a combination of these states. A significant perisurgical catecholamine surge produces unopposed sympathetic effects,12 increasing the risk of rupture of vulnerable coronary plaques, thrombus formation, and adverse cardiac events.13, 14 Similarly, augmented inflammatory responses and increased circulating coagulation factors further predispose to a hazardous perioperative milieu.15 Statins attenuate this vascular inflammatory response by suppressing the synthesis of mevalonate by inhibiting HMG‐CoA reductase. Suppression of mevalonate synthesis reduces the bioavailability of 2 important isoprenoid molecules: farnesyl‐pyrophosphate and geranylgeranyl‐pyrophosphate.16 Diminution of these isoprenoid intermediaries leads to reductions in the active intracellular signaling molecules Ras, Rho, and Rac, which play critical roles in vascular reactivity, endothelial function, and coagulation and inflammatory pathways.1723 The cumulative effect of these cellular changes is diminished inflammation during periods of surgical stress (Figure 1).

Figure 1

While the perioperative pleiotropicity of statins is of inherent clinical value, several studies have shown that these effects are lost and even reversed when statins are withdrawn.2428 During statin treatment, absence of isoprenoid intermediaries induces cytosolic accumulation of nonactivated Rho and Rac proteins. Abrupt cessation of statins activates Rho/Rac‐kinase pathways, leading to unregulated inflammation, platelet hyper‐activation, and endothelial dysfunction.24, 25, 28, 29 For instance, statin withdrawal in mice‐models leads to an overshoot activation of Rho, resulting in down‐regulation of endothelial nitric oxide production,25 activation of nicotinamide adenine dinucleotide phosphate (NAD[P]H)‐oxidase, and increased superoxide production.29 In another mouse‐model, statin withdrawal was associated with up‐regulation of key pro‐thrombotic molecules including platelet factor 4 and beta‐thromboglobulin.24 In human studies, a platelet hyper‐activation state (manifested by increased platelet P‐selectin expression and enhanced platelet aggregation) occurs after statin discontinuation.27 Furthermore, withdrawal of statins in patients with hyperlipidemia increases inflammatory markers such as C‐reactive protein and interleukin‐6.26 In the perioperative context, absence of these important anti‐inflammatory properties increases the risk of cardiac events.9, 10

EVIDENCE SUGGESTING BENEFIT FROM PERIOPERATIVE STATIN TREATMENT

Retrospective studies first suggested clinical benefit from perioperative statin treatment. In a case‐control study involving 2816 patients undergoing vascular surgery at Erasmus Medical Center, statin use was associated with substantially decreased postoperative mortality (adjusted odd ratio [OR] 0.22, 95% confidence interval [CI] 0.100.47).5 In a subsequent retrospective cohort study of 780,591 patients who underwent major noncardiac surgery, the risk of postoperative mortality was considerably lower among statin users (unadjusted OR 0.68, 95% CI 0.640.72) compared to patients who did not receive, or received delayed treatment with statins.3 A third retrospective study of 1163 vascular surgery patients found that statins prevented perioperative cardiac complications including death, MI, congestive heart failure, and ventricular tachyarrhythmias (OR 0.52, 95% CI 0.350.76).4

The benefit from statin treatment found in retrospective studies prompted the first double‐blinded, randomized controlled trial (RCT) of perioperative statin use. In 2004, Durazzo and colleagues1 randomized 100 statin‐naive patients scheduled to undergo elective aortic, femoro‐popliteal, or carotid surgery to receive either 20 mg of atorvastatin or placebo for 45 days. Vascular surgery was performed, on average, 31 days after randomization. Atorvastatin therapy reduced the incidence of death from cardiac causes, nonfatal acute MI, ischemic stroke, and unstable angina (26% in the placebo group vs 8% in the atorvastatin group; P = 0.031).1 Although the small size of the trial rendered it underpowered to show a mortality benefit, this remains the first RCT to demonstrate a protective perioperative effect of statins.

In the 2009 Dutch Echocardiographic Cardiac Risk Evaluation Applying Stress Echocardiography (DECREASE)‐III trial, Schouten and colleagues6 randomized 497 high‐risk, statin‐naive patients undergoing vascular surgery to receive, in addition to a beta‐blocker, either fluvastatin or placebo before surgery (median of 37 days). Postoperative myocardial ischemia (hazard ratio [HR] 0.55, 95% CI 0.340.88), and combined death from cardiovascular causes or nonfatal MI (HR 0.47, 95% CI 0.240.94), occurred less frequently in the treatment group.6 In 2009, the same group published DECREASE‐IV, a multicenter, prospective, open‐label, 2 2 factorial design trial of 1066 intermediate‐risk patients, scheduled to undergo elective, noncardiac surgery. Patients were assigned to bisoprolol, fluvastatin, combination treatment, or control therapy before surgery (median of 34 days). Although those randomized to fluvastatin demonstrated lower incidence of 30‐day cardiac death and MI than control (HR 0.65, 95% CI 0.351.10), these outcomes failed to reach statistical significance as the trial was principally powered to examine the effects of perioperative beta‐blockade.30

Using this pool of data, a meta‐analysis of 15 studies (223,010 patients) found a substantial 38% reduction in the risk of mortality after cardiac surgery (1.9% vs 3.1%; P = 0.0001) and an even greater 59% reduction in the risk of mortality following vascular surgery (1.7% vs 6.1%; P = 0.0001) with perioperative statin therapy. When including noncardiac surgery, a 44% reduction in mortality was observed (2.2% vs 3.2%; P < 0.01).2 We performed a similar meta‐analysis of 15 RCTs involving 2292 patients to determine whether perioperative statin treatment in statin‐naive patients, undergoing either cardiac or noncardiac surgery, improved clinical outcomes. Our analysis also found statistically significant reductions in the risk of MI associated with perioperative statin use in both cardiac and noncardiac surgery (risk reduction [RR] 0.53, 95% CI 0.380.74) and atrial fibrillation in statin‐naive patients undergoing cardiac surgery (RR 0.56, 95% CI 0.450.69).31 Taken together, a large volume of evidence supports the use of statins in surgical settings.

In view of this evidence, the ACCF/AHA perioperative guidelines for noncardiac surgery endorsed statins as an important risk‐reducing intervention in those undergoing noncardiac surgery, and recommended continued use in patients on chronic statin treatment scheduled for noncardiac surgery (Level of Evidence B, Class I; Benefits >>> Risk). Initiating statins in patients undergoing vascular surgery, with or without risk factors, was considered reasonable (Level of Evidence B, Class IIa; Benefits >> Risk).7 Current ESC perioperative guidelines in noncardiac surgery offer similar recommendations to those of ACCF/AHA, but differ by categorizing the recommendation to initiate statins in patients at high cardiovascular risk as a Class I recommendation.8

CLINICAL CONSEQUENCES OF STATIN WITHDRAWAL

Although statins provide important cardiac benefits, an important limitation to their perioperative use remains their oral‐only formulation. Thus, patients who are unable to resume oral intake may fail to resume treatment. Perioperative statin cessation has been hypothesized to lead to a statin withdrawal phenomenon. The evidence that supports the existence of this phenomenon comes from 3 distinct populations: ACS, ischemic stroke, and perioperative patients (Table 1).

CLINICAL INSIGHTS INTO FAILURE OF POSTOPERATIVE STATIN RESUMPTION

We hypothesize that failure to resume perioperative statins may occur for 4 cardinal reasons. First, resumption of an oral agent frequently proves clinically challenging when complications such as postoperative ileus, nausea, and vomiting peak. To date, no intravenous statin formulations are available, although phase‐I studies are currently underway.38 Second, it is not inconceivable that perioperative clinical instability may overshadow the resumption of statin treatment. Third, clinicians may also remain concerned regarding adverse effects of statins, a thought compounded by US Food and Drug Administration statin package inserts that specifically advocate for statins to be withheld during surgery. However, although the occurrence of elevated liver function tests and myopathy are theoretically important, the overwhelming majority of perioperative statin studies in noncardiac surgery have not found this to be a major occurrence.39 Nonetheless, a lack of uniform definitions and appropriate surveillance for adverse events are important limitations to this finding. In our recent systematic review, we were unable to provide refined estimates of these important side effects owing to differences in definition, variations in screening, and absence of standardized cutoffs used in studies.31 Finally, an important reason for failing to resume postoperative statins is that many physicians simply fail to recognize the perioperative importance of these agents.

STRATEGIES TO IMPROVE PERIOPERATIVE STATIN RESUMPTION

Using the existing evidence, we propose the following 4 clinical strategies to assist in avoiding a statin withdrawal state.

Preoperative Transition to Extended Release Statin Formulations

An innovative approach to minimizing statin withdrawal involves preoperative transition to an extended‐release statin formulation. This strategy may be of particular value in patients where prolonged bowel nonavailability is likely, such as those undergoing gastrointestinal surgery, or when prolonged postoperative dietary restriction (eg, NPO [nil per os]: nothing by mouth) status is expected (Figure 2).

Figure 2

CONCLUSIONS AND FUTURE DIRECTIONS

Sudden withdrawal of perioperative statins results in adverse clinical outcomes. Individuals engaged in the care of patients during surgery such as hospitalists, anesthesiologists, and surgeons must become more cognizant of a statin withdrawal state.

An important limitation associated with the study of perioperative statin withdrawal remains the ambiguity regarding the extent of the problem in the United States. Therefore, a logical first step could be the use of infrastructure within the National Surgical Quality Improvement Program (NSQIP) to understand the epidemiology of perioperative statin use and consequences associated with statin discontinuation.44 Mandating such quality reporting could easily be built into current NSQIP performance metrics. These data would help inform a research agenda targeting patients that experience statin withdrawal and strategies most likely to prevent it.

Healthcare market research has predicted that over 80% of physicians will use smartphones by 2012.1 These handheld devices allow users immediate access to various forms of electronic media such as Internet, instant messaging, and e‐mail. Smartphones provide numerous benefits to physicians, including rapid access to medical references, research applications, and patient information.2 These devices have been used for teleconsultation3 and patient education,4 and applications have been developed for numerous clinical specialties.48

Housestaff perceive that communication improves when they use smartphones rather than traditional pagers on the inpatient service,9 and patients may have a positive view of physicians' use of handheld computers.10 Medical schools and residency programs are increasingly requiring smartphone ownership for their trainees, with the expectation that smartphone use will enhance the educational experience, ensure the highest level of patient care, improve user efficiency, and help control the costs associated with purchasing updated textbooks.7, 1113 In the future, hospitals may rely on smartphone technologies to help reduce the enormous economic burden created by inefficient communication.14

Despite their numerous benefits for physicians and patients, little is known about the potential for smartphones to distract users in clinical care settings. Studies from the psychology and traffic safety fields have documented untoward consequences when individuals use electronic devices to multitask.1519 Given these concerns, we investigated the prevalence and patterns of smartphone use during inpatient attending rounds, and whether these devices can distract team members in this period of important information transfer.

METHODS

At our institution, attending rounds are faculty‐led inpatient teaching rounds that focus on clinical care and patient management; these sessions may be conducted either in the classroom or at the bedside, depending on patient and learner needs, and faculty preference. Inpatient teams are comprised of 1 attending, housestaff, and third and fourth year medical students. Each team conducts attending rounds independently; these rounds range in length from 1 hour (Pediatrics) to 2 hours (Medicine).

A survey instrument was designed to evaluate smartphone usage patterns during hospital inpatient attending rounds, and perceived distraction from smartphones in this setting. A preliminary version of the survey was pilot tested by a group of housestaff for face validity, redundancy, and ease of use, and it was subsequently revised. For the purposes of this study, a smartphone was defined broadly as any mobile, personal communication device (cellphone, iPhone, Android, Blackberry, iPad, etc). Residents were asked about their own smartphone use, as well as their observations of supervising attendings and other learners' devices use during rounds (see Supporting Appendix 1 in the online version of this article).

In February 2011, the anonymous online survey was administered using Survey Monkey (www.surveymonkey.com) and was distributed via e‐mail to medical and pediatric housestaff at Jacobi Medical Center, a public teaching hospital located in Bronx, NY, affiliated with the Albert Einstein College of Medicine of Yeshiva University. A similar survey was distributed to faculty who were known to conduct inpatient attending rounds (see Supporting Appendix 2 in the online version of this article). The e‐mail solicitation process was repeated for both groups of respondents at 2 and 4 weeks after the initial request. The study was approved by the Institutional Review Board of the Albert Einstein College of Medicine.

Respondents were not required to answer each question in order to complete the survey. With the exception of free‐text comments, all responses were either yes/no or were graded on a 5‐point frequency scale (1 = never, 2 = rarely, 3 = sometimes, 4 = often, 5 = always). This scale was chosen because it allowed for adequate dispersion of responses, and for the identification of meaningful smartphone usage among respondents (score 3) and data dichotomization. The z test was used to compare the proportions between independent groups.

All free‐text comments were imported into a Microsoft Word table. Comments were separated into 2 groups: housestaff and attending. Each comment was hand‐coded by 2 authors (R.J.K.‐S. and R.S.) to reach consensus for 1 of the following 4 categories: the comment was a positive statement; a negative statement; a positive/negative statement; or a neutral one, ie, neither positive nor negative. The terms positive and negative here refer to whether the statement explicitly highlighted benefits of smartphone use or a negative aspect of smartphone use, respectively. A comment was coded as positive/negative if it highlighted both benefits and drawbacks in the same comment. In addition, each comment that mentioned texting or call functions was secondarily coded as personal, patient, both, or unknown depending on the purpose of the texting or calls described in each comment. Comments were also reviewed for possible subthemes.

RESULTS

The overall response rate was 73% (156/214), with 81% (116/143) of housestaff and 56% (40/71) of faculty participating. The mean tenure of faculty respondents was 13 years. Eighty‐nine percent (103/116) of residents and 98% (39/40) of faculty owned devices, with 57% of housestaff and 28% of attendings reporting regular personal use of smartphones during attending rounds (Table 1).

Respondents reported that they used their smartphones during attending rounds for the following reasons: 1) patient care (85% residents, 48% faculty); 2) reading/responding to personal texts/e‐mails (37% residents, 12% faculty); and 3) other non‐patient care uses, such as Web surfing (15% residents, 0% faculty) (Tables 2 and 3). Nineteen percent of residents reported that they missed important clinical information due to distraction from smartphone use, as did 12% of attendings (Table 4). Respondents reported observing other team members using smartphones and missing important clinical data at higher rates than they reported for themselves (see Tables 1, 4, and 5). A majority of both residents (56%) and faculty (73%) agreed (score >3) that smartphones can be a serious distraction during attending rounds, and 77% of attendings affirmed that teaching hospitals should establish smartphone use codes of conduct in order to minimize unnecessary distraction during attending rounds.

Despite not requiring responses in order to complete the questionnaires, we found that, in general, few eligible faculty or residents skipped questions on the survey. Nevertheless, there was a substantial drop in responses (91/116) for the last 2 questions on the housestaff survey. These questions asked for resident observations of attending smartphone usage patterns during rounds, and whether they had seen attendings miss clinical information because of distractions from smartphone use.

There were 25 free‐text comments from residents and 11 from attendings. The resultant comments highlight differences in residents' and attendings' perspectives toward smartphone use during attending rounds. Housestaff comments included 7 positive comments, 7 positive/negative comments, 1 negative comment, and 10 neutral comments. A subtheme that emerged in 2 of the housestaff comments was the importance of personal autonomy in being able to use one's smartphone. Attending comments included 2 positive comments, 0 positive/negative comments, 4 negative comments, and 5 neutral comments. Faculty comments revealed that attendings use their smartphones' e‐mail/texting and call capabilities during rounds both for patient care issues (3 comments) and/or urgent family concerns (2 comments). In 2 other attending comments, the reason for calls/texts during rounds was not specified.

Housestaff comments included: I do not know why it is that attendings never use it these phones are so easy to use and [enhance] patient care in a number of ways, Depending on how they are used, if strictly for pt care then they can be a great mobile tool, Of course they can be a distraction, but they are also a very good tool. You take the good with the bad, If you are bored you will find other things to occupy your mind. If you can look up some info at the time of rounding you are actively participating. Please, do not make it worse than it is already, and It is a personal choice. Faculty negative comments highlighted the potential for distraction from the e‐mail beeps, the fact that some of the housestaff will be tuned into their SmartPhones, that residents frequently check their phones during roundsa distraction and frankly rude when the attending or fellow are giving a brief lecture, and that sometimes more focus is on the SmartPhone than rounds.

DISCUSSION

Physicians and their patients benefit from the wide‐ranging capabilities of personal, mobile communication devices in the healthcare environment. Smartphones house the latest medical references, provide access to patients' medical records and imaging studies, can photograph or video physical findings, and educate and monitor patients.28 Smartphones can facilitate information transfer in the medical setting and may improve housestaff efficiency and communication.9

Despite their significant benefits, smartphones introduce another source of interruption, multitasking, and distraction into the hospital environment. There is increasing awareness that breaks‐in‐task in the clinical setting may have negative consequences.2024 While some types of interruptions are beneficial and can facilitate patient care (eg, an alarm ringing to indicate abnormal vitals signs on a patient),2024 other forms of interruptions, even those that are self‐initiated,22 can be distracting and detrimental. Along these lines, recommendations for safe handoffs and information transfer have specifically included advice to minimize potential distractions.25

In addition, studies from the psychology and education literature have previously documented negative consequences on learning when individuals use electronic devices to multitask.1517 Students who used a laptop in class were likely to multitask, become distracted, and distract others; the more a student used the laptop in class, the lower the student's class performance.15 Multitasking with a cellphone during driving can be especially hazardous.18, 19 According to National Highway Traffic Safety Administration data, 20% of injury crashes in 2009 involved reports of distracted driving, and cellphones were implicated in 18% of distracted driving deaths that year.18

Little is known about any negative effects of using personal electronic devices in the context of patient care. A 2011 study of Internal Medicine residents who used smartphones for team communication documented both positive and negative consequences of smartphone use in the hospital setting. Negative consequences included frequent interruptions, a weakening of interprofessional behaviors as housestaff relied on texting over direct communication with nurses, and unprofessional housestaff behaviors.26 The Agency for Healthcare Quality and Research published a case report in which a resident's smartphone use during clinical care resulted in patient harm.27 To our knowledge, this is the first study to detail housestaff and faculty smartphone usage patterns and potential for user distractibility during inpatient attending rounds.

Our data show that device use during attending rounds is prevalent among residents and faculty alike, with the majority of use related to patient care. However, attendings were half as likely as residents to report using devices regularly during rounds. This finding may reflect attendings' inability to multitask while leading the rounds, or a deliberate role‐modeling of desired conduct during rounds. Generational differences may also play a role, with residents more likely than their older attendings to multitask and self‐interrupt. Along these lines, traffic safety research has found that younger drivers are more likely to text during driving; approximately 30% of drivers under 30 years old reported texting while driving in the previous 30 days, compared to 9% of respondents over 30 years old.19 Increased smartphone use by housestaff during rounds may also reflect attitudinal differences between the 2 groups. As seen in the free‐text comments, housestaff tended to emphasize the benefits of smartphone use, and with 1 exception, all negative housestaff comments were balanced by a positive statement. Faculty more commonly underscored the negative aspects of smartphone use during rounds, including the devices' adverse effects on housestaff professional behavior in this setting.

Faculty and housestaff consistently reported observing others using smartphones at higher rates than they reported for themselves. This discrepancy may reflect underrecognition of self‐use, or a discomfort in reporting self‐use during attending rounds. In addition, residents' observations of other trainees' usage of smartphones (91%) was higher than faculty observation of the same group (73%). Trainees' smartphone use may be less obvious to attendings who are involved in facilitating rounds. Alternatively, trainees may use their smartphones in subtle ways to prevent attending awareness.

There are several limitations to our study. Our research focused specifically on attending rounds. Smartphone usage patterns by faculty and housestaff at other times in the work day, such as during resident handoffs, at a patient's bedside, or during academic conferences, may differ. Nevertheless, we specifically chose to study smartphone use during attending rounds, as these sessions are discrete time frames during which important teaching occurs and clinical management decisions are made. With recent Accreditation Council for Graduate Medical Education (ACGME) work hour restrictions, these faculty‐led rounds may become increasingly important in ensuring the safe transition of patient care. Secondly, despite asking respondents how often they use their smartphones for personal texts or e‐mails, it was clear from the free‐text comments that respondents use their smartphone e‐mail/texting capabilities and take urgent calls during rounds for both patient care and/or family issues. It is not possible from the data to sort out the subset of respondents who use texting or e‐mailing exclusively for patient care during rounds. Third, we did not survey medical students on the teams, so it is possible that their device use on rounds differs from that of housestaff and faculty. Fourth, since the survey could be completed without answering every question, response rates for some items varied slightly; there was a substantial reduction in the number of eligible residents who answered the final 2 questions on the survey about their observations of attendings' smartphone usage patterns and distraction during rounds. While the flexibility in survey completion was intended to enhance overall study participation, it is unknown how nonresponders might have affected the study results; as such, those specific results should be interpreted with some caution. Finally, our findings were based on respondents' retrospective recall, and therefore may not accurately reflect true usage patterns. Timemotion studies with real‐time observation of smartphone use would provide more accurate data.

A majority of residents and attendings in our study agreed that smartphones can pose a serious distraction during attending rounds, and attendings strongly favored the institution of formal codes of conduct for smartphone use during inpatient attending rounds. The development of such policies are important for patient safety; at the same time, they are in line with medical institutions' increasing awareness about the need for guidelines regarding other aspects of digital professionalism.28 In February 2012, our hospital instituted a policy regarding appropriate device use during inpatient attending rounds (see Supporting Appendix 3 in the online version of this article). Because our research found differences in housestaff and faculty attitudes toward smartphone use during rounds, we developed our policy after discussion with, and feedback from, all members of the inpatient team, including faculty, residents, and medical students. Incorporating the various perspectives of all stakeholders can be helpful to institutions in developing guidelines that maximize the benefits of smartphone use in the learning environment, while reducing the potential for distraction and adverse outcomes.

Healthcare market research has predicted that over 80% of physicians will use smartphones by 2012.1 These handheld devices allow users immediate access to various forms of electronic media such as Internet, instant messaging, and e‐mail. Smartphones provide numerous benefits to physicians, including rapid access to medical references, research applications, and patient information.2 These devices have been used for teleconsultation3 and patient education,4 and applications have been developed for numerous clinical specialties.48

Housestaff perceive that communication improves when they use smartphones rather than traditional pagers on the inpatient service,9 and patients may have a positive view of physicians' use of handheld computers.10 Medical schools and residency programs are increasingly requiring smartphone ownership for their trainees, with the expectation that smartphone use will enhance the educational experience, ensure the highest level of patient care, improve user efficiency, and help control the costs associated with purchasing updated textbooks.7, 1113 In the future, hospitals may rely on smartphone technologies to help reduce the enormous economic burden created by inefficient communication.14

Despite their numerous benefits for physicians and patients, little is known about the potential for smartphones to distract users in clinical care settings. Studies from the psychology and traffic safety fields have documented untoward consequences when individuals use electronic devices to multitask.1519 Given these concerns, we investigated the prevalence and patterns of smartphone use during inpatient attending rounds, and whether these devices can distract team members in this period of important information transfer.

METHODS

At our institution, attending rounds are faculty‐led inpatient teaching rounds that focus on clinical care and patient management; these sessions may be conducted either in the classroom or at the bedside, depending on patient and learner needs, and faculty preference. Inpatient teams are comprised of 1 attending, housestaff, and third and fourth year medical students. Each team conducts attending rounds independently; these rounds range in length from 1 hour (Pediatrics) to 2 hours (Medicine).

A survey instrument was designed to evaluate smartphone usage patterns during hospital inpatient attending rounds, and perceived distraction from smartphones in this setting. A preliminary version of the survey was pilot tested by a group of housestaff for face validity, redundancy, and ease of use, and it was subsequently revised. For the purposes of this study, a smartphone was defined broadly as any mobile, personal communication device (cellphone, iPhone, Android, Blackberry, iPad, etc). Residents were asked about their own smartphone use, as well as their observations of supervising attendings and other learners' devices use during rounds (see Supporting Appendix 1 in the online version of this article).

In February 2011, the anonymous online survey was administered using Survey Monkey (www.surveymonkey.com) and was distributed via e‐mail to medical and pediatric housestaff at Jacobi Medical Center, a public teaching hospital located in Bronx, NY, affiliated with the Albert Einstein College of Medicine of Yeshiva University. A similar survey was distributed to faculty who were known to conduct inpatient attending rounds (see Supporting Appendix 2 in the online version of this article). The e‐mail solicitation process was repeated for both groups of respondents at 2 and 4 weeks after the initial request. The study was approved by the Institutional Review Board of the Albert Einstein College of Medicine.

Respondents were not required to answer each question in order to complete the survey. With the exception of free‐text comments, all responses were either yes/no or were graded on a 5‐point frequency scale (1 = never, 2 = rarely, 3 = sometimes, 4 = often, 5 = always). This scale was chosen because it allowed for adequate dispersion of responses, and for the identification of meaningful smartphone usage among respondents (score 3) and data dichotomization. The z test was used to compare the proportions between independent groups.

All free‐text comments were imported into a Microsoft Word table. Comments were separated into 2 groups: housestaff and attending. Each comment was hand‐coded by 2 authors (R.J.K.‐S. and R.S.) to reach consensus for 1 of the following 4 categories: the comment was a positive statement; a negative statement; a positive/negative statement; or a neutral one, ie, neither positive nor negative. The terms positive and negative here refer to whether the statement explicitly highlighted benefits of smartphone use or a negative aspect of smartphone use, respectively. A comment was coded as positive/negative if it highlighted both benefits and drawbacks in the same comment. In addition, each comment that mentioned texting or call functions was secondarily coded as personal, patient, both, or unknown depending on the purpose of the texting or calls described in each comment. Comments were also reviewed for possible subthemes.

RESULTS

The overall response rate was 73% (156/214), with 81% (116/143) of housestaff and 56% (40/71) of faculty participating. The mean tenure of faculty respondents was 13 years. Eighty‐nine percent (103/116) of residents and 98% (39/40) of faculty owned devices, with 57% of housestaff and 28% of attendings reporting regular personal use of smartphones during attending rounds (Table 1).

Respondents reported that they used their smartphones during attending rounds for the following reasons: 1) patient care (85% residents, 48% faculty); 2) reading/responding to personal texts/e‐mails (37% residents, 12% faculty); and 3) other non‐patient care uses, such as Web surfing (15% residents, 0% faculty) (Tables 2 and 3). Nineteen percent of residents reported that they missed important clinical information due to distraction from smartphone use, as did 12% of attendings (Table 4). Respondents reported observing other team members using smartphones and missing important clinical data at higher rates than they reported for themselves (see Tables 1, 4, and 5). A majority of both residents (56%) and faculty (73%) agreed (score >3) that smartphones can be a serious distraction during attending rounds, and 77% of attendings affirmed that teaching hospitals should establish smartphone use codes of conduct in order to minimize unnecessary distraction during attending rounds.

Despite not requiring responses in order to complete the questionnaires, we found that, in general, few eligible faculty or residents skipped questions on the survey. Nevertheless, there was a substantial drop in responses (91/116) for the last 2 questions on the housestaff survey. These questions asked for resident observations of attending smartphone usage patterns during rounds, and whether they had seen attendings miss clinical information because of distractions from smartphone use.

There were 25 free‐text comments from residents and 11 from attendings. The resultant comments highlight differences in residents' and attendings' perspectives toward smartphone use during attending rounds. Housestaff comments included 7 positive comments, 7 positive/negative comments, 1 negative comment, and 10 neutral comments. A subtheme that emerged in 2 of the housestaff comments was the importance of personal autonomy in being able to use one's smartphone. Attending comments included 2 positive comments, 0 positive/negative comments, 4 negative comments, and 5 neutral comments. Faculty comments revealed that attendings use their smartphones' e‐mail/texting and call capabilities during rounds both for patient care issues (3 comments) and/or urgent family concerns (2 comments). In 2 other attending comments, the reason for calls/texts during rounds was not specified.

Housestaff comments included: I do not know why it is that attendings never use it these phones are so easy to use and [enhance] patient care in a number of ways, Depending on how they are used, if strictly for pt care then they can be a great mobile tool, Of course they can be a distraction, but they are also a very good tool. You take the good with the bad, If you are bored you will find other things to occupy your mind. If you can look up some info at the time of rounding you are actively participating. Please, do not make it worse than it is already, and It is a personal choice. Faculty negative comments highlighted the potential for distraction from the e‐mail beeps, the fact that some of the housestaff will be tuned into their SmartPhones, that residents frequently check their phones during roundsa distraction and frankly rude when the attending or fellow are giving a brief lecture, and that sometimes more focus is on the SmartPhone than rounds.

DISCUSSION

Physicians and their patients benefit from the wide‐ranging capabilities of personal, mobile communication devices in the healthcare environment. Smartphones house the latest medical references, provide access to patients' medical records and imaging studies, can photograph or video physical findings, and educate and monitor patients.28 Smartphones can facilitate information transfer in the medical setting and may improve housestaff efficiency and communication.9

Despite their significant benefits, smartphones introduce another source of interruption, multitasking, and distraction into the hospital environment. There is increasing awareness that breaks‐in‐task in the clinical setting may have negative consequences.2024 While some types of interruptions are beneficial and can facilitate patient care (eg, an alarm ringing to indicate abnormal vitals signs on a patient),2024 other forms of interruptions, even those that are self‐initiated,22 can be distracting and detrimental. Along these lines, recommendations for safe handoffs and information transfer have specifically included advice to minimize potential distractions.25

In addition, studies from the psychology and education literature have previously documented negative consequences on learning when individuals use electronic devices to multitask.1517 Students who used a laptop in class were likely to multitask, become distracted, and distract others; the more a student used the laptop in class, the lower the student's class performance.15 Multitasking with a cellphone during driving can be especially hazardous.18, 19 According to National Highway Traffic Safety Administration data, 20% of injury crashes in 2009 involved reports of distracted driving, and cellphones were implicated in 18% of distracted driving deaths that year.18

Little is known about any negative effects of using personal electronic devices in the context of patient care. A 2011 study of Internal Medicine residents who used smartphones for team communication documented both positive and negative consequences of smartphone use in the hospital setting. Negative consequences included frequent interruptions, a weakening of interprofessional behaviors as housestaff relied on texting over direct communication with nurses, and unprofessional housestaff behaviors.26 The Agency for Healthcare Quality and Research published a case report in which a resident's smartphone use during clinical care resulted in patient harm.27 To our knowledge, this is the first study to detail housestaff and faculty smartphone usage patterns and potential for user distractibility during inpatient attending rounds.

Our data show that device use during attending rounds is prevalent among residents and faculty alike, with the majority of use related to patient care. However, attendings were half as likely as residents to report using devices regularly during rounds. This finding may reflect attendings' inability to multitask while leading the rounds, or a deliberate role‐modeling of desired conduct during rounds. Generational differences may also play a role, with residents more likely than their older attendings to multitask and self‐interrupt. Along these lines, traffic safety research has found that younger drivers are more likely to text during driving; approximately 30% of drivers under 30 years old reported texting while driving in the previous 30 days, compared to 9% of respondents over 30 years old.19 Increased smartphone use by housestaff during rounds may also reflect attitudinal differences between the 2 groups. As seen in the free‐text comments, housestaff tended to emphasize the benefits of smartphone use, and with 1 exception, all negative housestaff comments were balanced by a positive statement. Faculty more commonly underscored the negative aspects of smartphone use during rounds, including the devices' adverse effects on housestaff professional behavior in this setting.

Faculty and housestaff consistently reported observing others using smartphones at higher rates than they reported for themselves. This discrepancy may reflect underrecognition of self‐use, or a discomfort in reporting self‐use during attending rounds. In addition, residents' observations of other trainees' usage of smartphones (91%) was higher than faculty observation of the same group (73%). Trainees' smartphone use may be less obvious to attendings who are involved in facilitating rounds. Alternatively, trainees may use their smartphones in subtle ways to prevent attending awareness.

There are several limitations to our study. Our research focused specifically on attending rounds. Smartphone usage patterns by faculty and housestaff at other times in the work day, such as during resident handoffs, at a patient's bedside, or during academic conferences, may differ. Nevertheless, we specifically chose to study smartphone use during attending rounds, as these sessions are discrete time frames during which important teaching occurs and clinical management decisions are made. With recent Accreditation Council for Graduate Medical Education (ACGME) work hour restrictions, these faculty‐led rounds may become increasingly important in ensuring the safe transition of patient care. Secondly, despite asking respondents how often they use their smartphones for personal texts or e‐mails, it was clear from the free‐text comments that respondents use their smartphone e‐mail/texting capabilities and take urgent calls during rounds for both patient care and/or family issues. It is not possible from the data to sort out the subset of respondents who use texting or e‐mailing exclusively for patient care during rounds. Third, we did not survey medical students on the teams, so it is possible that their device use on rounds differs from that of housestaff and faculty. Fourth, since the survey could be completed without answering every question, response rates for some items varied slightly; there was a substantial reduction in the number of eligible residents who answered the final 2 questions on the survey about their observations of attendings' smartphone usage patterns and distraction during rounds. While the flexibility in survey completion was intended to enhance overall study participation, it is unknown how nonresponders might have affected the study results; as such, those specific results should be interpreted with some caution. Finally, our findings were based on respondents' retrospective recall, and therefore may not accurately reflect true usage patterns. Timemotion studies with real‐time observation of smartphone use would provide more accurate data.

A majority of residents and attendings in our study agreed that smartphones can pose a serious distraction during attending rounds, and attendings strongly favored the institution of formal codes of conduct for smartphone use during inpatient attending rounds. The development of such policies are important for patient safety; at the same time, they are in line with medical institutions' increasing awareness about the need for guidelines regarding other aspects of digital professionalism.28 In February 2012, our hospital instituted a policy regarding appropriate device use during inpatient attending rounds (see Supporting Appendix 3 in the online version of this article). Because our research found differences in housestaff and faculty attitudes toward smartphone use during rounds, we developed our policy after discussion with, and feedback from, all members of the inpatient team, including faculty, residents, and medical students. Incorporating the various perspectives of all stakeholders can be helpful to institutions in developing guidelines that maximize the benefits of smartphone use in the learning environment, while reducing the potential for distraction and adverse outcomes.

Healthcare market research has predicted that over 80% of physicians will use smartphones by 2012.1 These handheld devices allow users immediate access to various forms of electronic media such as Internet, instant messaging, and e‐mail. Smartphones provide numerous benefits to physicians, including rapid access to medical references, research applications, and patient information.2 These devices have been used for teleconsultation3 and patient education,4 and applications have been developed for numerous clinical specialties.48

Housestaff perceive that communication improves when they use smartphones rather than traditional pagers on the inpatient service,9 and patients may have a positive view of physicians' use of handheld computers.10 Medical schools and residency programs are increasingly requiring smartphone ownership for their trainees, with the expectation that smartphone use will enhance the educational experience, ensure the highest level of patient care, improve user efficiency, and help control the costs associated with purchasing updated textbooks.7, 1113 In the future, hospitals may rely on smartphone technologies to help reduce the enormous economic burden created by inefficient communication.14

Despite their numerous benefits for physicians and patients, little is known about the potential for smartphones to distract users in clinical care settings. Studies from the psychology and traffic safety fields have documented untoward consequences when individuals use electronic devices to multitask.1519 Given these concerns, we investigated the prevalence and patterns of smartphone use during inpatient attending rounds, and whether these devices can distract team members in this period of important information transfer.

METHODS

At our institution, attending rounds are faculty‐led inpatient teaching rounds that focus on clinical care and patient management; these sessions may be conducted either in the classroom or at the bedside, depending on patient and learner needs, and faculty preference. Inpatient teams are comprised of 1 attending, housestaff, and third and fourth year medical students. Each team conducts attending rounds independently; these rounds range in length from 1 hour (Pediatrics) to 2 hours (Medicine).

A survey instrument was designed to evaluate smartphone usage patterns during hospital inpatient attending rounds, and perceived distraction from smartphones in this setting. A preliminary version of the survey was pilot tested by a group of housestaff for face validity, redundancy, and ease of use, and it was subsequently revised. For the purposes of this study, a smartphone was defined broadly as any mobile, personal communication device (cellphone, iPhone, Android, Blackberry, iPad, etc). Residents were asked about their own smartphone use, as well as their observations of supervising attendings and other learners' devices use during rounds (see Supporting Appendix 1 in the online version of this article).

In February 2011, the anonymous online survey was administered using Survey Monkey (www.surveymonkey.com) and was distributed via e‐mail to medical and pediatric housestaff at Jacobi Medical Center, a public teaching hospital located in Bronx, NY, affiliated with the Albert Einstein College of Medicine of Yeshiva University. A similar survey was distributed to faculty who were known to conduct inpatient attending rounds (see Supporting Appendix 2 in the online version of this article). The e‐mail solicitation process was repeated for both groups of respondents at 2 and 4 weeks after the initial request. The study was approved by the Institutional Review Board of the Albert Einstein College of Medicine.

Respondents were not required to answer each question in order to complete the survey. With the exception of free‐text comments, all responses were either yes/no or were graded on a 5‐point frequency scale (1 = never, 2 = rarely, 3 = sometimes, 4 = often, 5 = always). This scale was chosen because it allowed for adequate dispersion of responses, and for the identification of meaningful smartphone usage among respondents (score 3) and data dichotomization. The z test was used to compare the proportions between independent groups.

All free‐text comments were imported into a Microsoft Word table. Comments were separated into 2 groups: housestaff and attending. Each comment was hand‐coded by 2 authors (R.J.K.‐S. and R.S.) to reach consensus for 1 of the following 4 categories: the comment was a positive statement; a negative statement; a positive/negative statement; or a neutral one, ie, neither positive nor negative. The terms positive and negative here refer to whether the statement explicitly highlighted benefits of smartphone use or a negative aspect of smartphone use, respectively. A comment was coded as positive/negative if it highlighted both benefits and drawbacks in the same comment. In addition, each comment that mentioned texting or call functions was secondarily coded as personal, patient, both, or unknown depending on the purpose of the texting or calls described in each comment. Comments were also reviewed for possible subthemes.

RESULTS

The overall response rate was 73% (156/214), with 81% (116/143) of housestaff and 56% (40/71) of faculty participating. The mean tenure of faculty respondents was 13 years. Eighty‐nine percent (103/116) of residents and 98% (39/40) of faculty owned devices, with 57% of housestaff and 28% of attendings reporting regular personal use of smartphones during attending rounds (Table 1).

Respondents reported that they used their smartphones during attending rounds for the following reasons: 1) patient care (85% residents, 48% faculty); 2) reading/responding to personal texts/e‐mails (37% residents, 12% faculty); and 3) other non‐patient care uses, such as Web surfing (15% residents, 0% faculty) (Tables 2 and 3). Nineteen percent of residents reported that they missed important clinical information due to distraction from smartphone use, as did 12% of attendings (Table 4). Respondents reported observing other team members using smartphones and missing important clinical data at higher rates than they reported for themselves (see Tables 1, 4, and 5). A majority of both residents (56%) and faculty (73%) agreed (score >3) that smartphones can be a serious distraction during attending rounds, and 77% of attendings affirmed that teaching hospitals should establish smartphone use codes of conduct in order to minimize unnecessary distraction during attending rounds.

Despite not requiring responses in order to complete the questionnaires, we found that, in general, few eligible faculty or residents skipped questions on the survey. Nevertheless, there was a substantial drop in responses (91/116) for the last 2 questions on the housestaff survey. These questions asked for resident observations of attending smartphone usage patterns during rounds, and whether they had seen attendings miss clinical information because of distractions from smartphone use.

There were 25 free‐text comments from residents and 11 from attendings. The resultant comments highlight differences in residents' and attendings' perspectives toward smartphone use during attending rounds. Housestaff comments included 7 positive comments, 7 positive/negative comments, 1 negative comment, and 10 neutral comments. A subtheme that emerged in 2 of the housestaff comments was the importance of personal autonomy in being able to use one's smartphone. Attending comments included 2 positive comments, 0 positive/negative comments, 4 negative comments, and 5 neutral comments. Faculty comments revealed that attendings use their smartphones' e‐mail/texting and call capabilities during rounds both for patient care issues (3 comments) and/or urgent family concerns (2 comments). In 2 other attending comments, the reason for calls/texts during rounds was not specified.

Housestaff comments included: I do not know why it is that attendings never use it these phones are so easy to use and [enhance] patient care in a number of ways, Depending on how they are used, if strictly for pt care then they can be a great mobile tool, Of course they can be a distraction, but they are also a very good tool. You take the good with the bad, If you are bored you will find other things to occupy your mind. If you can look up some info at the time of rounding you are actively participating. Please, do not make it worse than it is already, and It is a personal choice. Faculty negative comments highlighted the potential for distraction from the e‐mail beeps, the fact that some of the housestaff will be tuned into their SmartPhones, that residents frequently check their phones during roundsa distraction and frankly rude when the attending or fellow are giving a brief lecture, and that sometimes more focus is on the SmartPhone than rounds.

DISCUSSION

Physicians and their patients benefit from the wide‐ranging capabilities of personal, mobile communication devices in the healthcare environment. Smartphones house the latest medical references, provide access to patients' medical records and imaging studies, can photograph or video physical findings, and educate and monitor patients.28 Smartphones can facilitate information transfer in the medical setting and may improve housestaff efficiency and communication.9

Despite their significant benefits, smartphones introduce another source of interruption, multitasking, and distraction into the hospital environment. There is increasing awareness that breaks‐in‐task in the clinical setting may have negative consequences.2024 While some types of interruptions are beneficial and can facilitate patient care (eg, an alarm ringing to indicate abnormal vitals signs on a patient),2024 other forms of interruptions, even those that are self‐initiated,22 can be distracting and detrimental. Along these lines, recommendations for safe handoffs and information transfer have specifically included advice to minimize potential distractions.25

In addition, studies from the psychology and education literature have previously documented negative consequences on learning when individuals use electronic devices to multitask.1517 Students who used a laptop in class were likely to multitask, become distracted, and distract others; the more a student used the laptop in class, the lower the student's class performance.15 Multitasking with a cellphone during driving can be especially hazardous.18, 19 According to National Highway Traffic Safety Administration data, 20% of injury crashes in 2009 involved reports of distracted driving, and cellphones were implicated in 18% of distracted driving deaths that year.18

Little is known about any negative effects of using personal electronic devices in the context of patient care. A 2011 study of Internal Medicine residents who used smartphones for team communication documented both positive and negative consequences of smartphone use in the hospital setting. Negative consequences included frequent interruptions, a weakening of interprofessional behaviors as housestaff relied on texting over direct communication with nurses, and unprofessional housestaff behaviors.26 The Agency for Healthcare Quality and Research published a case report in which a resident's smartphone use during clinical care resulted in patient harm.27 To our knowledge, this is the first study to detail housestaff and faculty smartphone usage patterns and potential for user distractibility during inpatient attending rounds.

Our data show that device use during attending rounds is prevalent among residents and faculty alike, with the majority of use related to patient care. However, attendings were half as likely as residents to report using devices regularly during rounds. This finding may reflect attendings' inability to multitask while leading the rounds, or a deliberate role‐modeling of desired conduct during rounds. Generational differences may also play a role, with residents more likely than their older attendings to multitask and self‐interrupt. Along these lines, traffic safety research has found that younger drivers are more likely to text during driving; approximately 30% of drivers under 30 years old reported texting while driving in the previous 30 days, compared to 9% of respondents over 30 years old.19 Increased smartphone use by housestaff during rounds may also reflect attitudinal differences between the 2 groups. As seen in the free‐text comments, housestaff tended to emphasize the benefits of smartphone use, and with 1 exception, all negative housestaff comments were balanced by a positive statement. Faculty more commonly underscored the negative aspects of smartphone use during rounds, including the devices' adverse effects on housestaff professional behavior in this setting.

Faculty and housestaff consistently reported observing others using smartphones at higher rates than they reported for themselves. This discrepancy may reflect underrecognition of self‐use, or a discomfort in reporting self‐use during attending rounds. In addition, residents' observations of other trainees' usage of smartphones (91%) was higher than faculty observation of the same group (73%). Trainees' smartphone use may be less obvious to attendings who are involved in facilitating rounds. Alternatively, trainees may use their smartphones in subtle ways to prevent attending awareness.

There are several limitations to our study. Our research focused specifically on attending rounds. Smartphone usage patterns by faculty and housestaff at other times in the work day, such as during resident handoffs, at a patient's bedside, or during academic conferences, may differ. Nevertheless, we specifically chose to study smartphone use during attending rounds, as these sessions are discrete time frames during which important teaching occurs and clinical management decisions are made. With recent Accreditation Council for Graduate Medical Education (ACGME) work hour restrictions, these faculty‐led rounds may become increasingly important in ensuring the safe transition of patient care. Secondly, despite asking respondents how often they use their smartphones for personal texts or e‐mails, it was clear from the free‐text comments that respondents use their smartphone e‐mail/texting capabilities and take urgent calls during rounds for both patient care and/or family issues. It is not possible from the data to sort out the subset of respondents who use texting or e‐mailing exclusively for patient care during rounds. Third, we did not survey medical students on the teams, so it is possible that their device use on rounds differs from that of housestaff and faculty. Fourth, since the survey could be completed without answering every question, response rates for some items varied slightly; there was a substantial reduction in the number of eligible residents who answered the final 2 questions on the survey about their observations of attendings' smartphone usage patterns and distraction during rounds. While the flexibility in survey completion was intended to enhance overall study participation, it is unknown how nonresponders might have affected the study results; as such, those specific results should be interpreted with some caution. Finally, our findings were based on respondents' retrospective recall, and therefore may not accurately reflect true usage patterns. Timemotion studies with real‐time observation of smartphone use would provide more accurate data.

A majority of residents and attendings in our study agreed that smartphones can pose a serious distraction during attending rounds, and attendings strongly favored the institution of formal codes of conduct for smartphone use during inpatient attending rounds. The development of such policies are important for patient safety; at the same time, they are in line with medical institutions' increasing awareness about the need for guidelines regarding other aspects of digital professionalism.28 In February 2012, our hospital instituted a policy regarding appropriate device use during inpatient attending rounds (see Supporting Appendix 3 in the online version of this article). Because our research found differences in housestaff and faculty attitudes toward smartphone use during rounds, we developed our policy after discussion with, and feedback from, all members of the inpatient team, including faculty, residents, and medical students. Incorporating the various perspectives of all stakeholders can be helpful to institutions in developing guidelines that maximize the benefits of smartphone use in the learning environment, while reducing the potential for distraction and adverse outcomes.