Advertisement

Characteristics and Outcomes of Suspected Digoxin Toxicity and Immune Fab Treatment Over the Past Two Decades—2000-2020

Open AccessPublished:September 09, 2022DOI:https://doi.org/10.1016/j.amjcard.2022.08.004
      The role of digoxin in clinical practice has narrowed over time. Data on digoxin toxicity trends and outcomes are variable and lack granularity for treatment outcomes. This study aimed to address data gaps in digoxin toxicity trends and outcomes in patients treated with or without digoxin immune fab (DIF). This single-center analysis examined patients with signs/symptoms concerning digoxin toxicity, defined as hospital admission or emergency department visit with elevated digoxin serum concentrations (>2 ng/ml) and/or a primary diagnosis code of digoxin toxicity and/or DIF order. Between 2000 and 2020, 727 patients were identified with signs concerning for digoxin toxicity with a mortality rate of 12.7% during admission and 42.7% at 1 year. DIF was ordered in 9% of cases. Incidence of digoxin toxicity per 1,000 patients with a digoxin prescription and frequency of DIF treatment fluctuated over time without a clear trend toward increase or reduction. DIF-treated patients demonstrated a heavier co-morbidity burden and lower presenting heart rates (median 53 [39.5 to 69.5] vs 77 [64.0 to 91.5] beats/min, p <0.001), worse renal function (median estimated glomerular filtration rate, 30.3 [14.8 to 48.6] vs 40.0 [24.2 to 61.2] ml/min/1.73 m2, p = 0.013), and higher potassium (median 4.5 [4.0 to 5.3] vs 4.3 [3.9 to 4.8] mEq/L, p = 0.022). Compared with a matched cohort, DIF-treated patients experienced a nonsignificant, numerically lower in-hospital mortality (8.2% vs 15.8%, p = 0.199) and 30-day all-cause hospitalization (14.3% vs 24.7%, p = 0.112) and similar 6-month and 1-year hospitalization and mortality. In conclusion, digoxin toxicity remains a pertinent public health issue despite reduction in digoxin utilization. DIF therapy is used in a medically complex population with a high-acuity illness at presentation and is associated with nonsignificant trends toward reduced in-hospital mortality and early readmission that are attenuated over time.

      Introduction

      The scope of use of digoxin in cardiovascular medicine has narrowed over time; in this setting, digoxin utilization has declined in the United States.
      • Haynes K
      • Heitjan DF
      • Kanetsky PA
      • Hennessy S.
      Declining public health burden of digoxin toxicity from 1991 to 2004.
      • Ambrosy AP
      • Bhatt AS
      • Stebbins AL
      • Wruck LM
      • Fudim M
      • Greene SJ
      • Kraus WE
      • O'Connor CMO
      • Piña IL
      • Whellan DJ
      • Mentz RJ
      Prevalent digoxin use and subsequent risk of death or hospitalization in ambulatory heart failure patients with a reduced ejection fraction - findings from the Heart Failure: a Controlled Trial Investigating Outcomes of Exercise Training (HF-ACTION) randomized controlled trial.
      • Angraal S
      • Nuti SV
      • Masoudi FA
      • Freeman JV
      • Murugiah K
      • Shah ND
      • Desai NR
      • Ranasinghe I
      • Wang Y
      • Krumholz HM.
      Digoxin use and associated adverse events among older adults.
      However, digoxin prescriptions remain reasonably prevalent (>4 million prescriptions dispensed by most recent data in 2014), and the drug maintains a select role, particularly in patients with heart failure (HF) and atrial fibrillation (AF).
      • Angraal S
      • Nuti SV
      • Masoudi FA
      • Freeman JV
      • Murugiah K
      • Shah ND
      • Desai NR
      • Ranasinghe I
      • Wang Y
      • Krumholz HM.
      Digoxin use and associated adverse events among older adults.
      ,
      • Kotecha D
      • Bunting KV
      • Gill SK
      • Mehta S
      • Stanbury M
      • Jones JC
      • Haynes S
      • Calvert MJ
      • Deeks JJ
      • Steeds RP
      • Strauss VY
      • Rahimi K
      • Camm AJ
      • Griffith M
      • Lip GYH
      • Townend JN
      • Kirchhof P.
      Rate Control Therapy Evaluation in Permanent Atrial Fibrillation (RATE-AF) Team. Effect of digoxin vs bisoprolol for heart rate control in atrial fibrillation on patient-reported quality of life: the RATE-AF randomized clinical trial.
      Owing to continued digoxin use, potential digoxin toxicity remains a common and serious public health problem. Although acute toxicity is uncommon, chronic digoxin toxicity has a relatively high incidence given the drug's narrow therapeutic window. Overall, rates of toxicity appear to be downtrending over time, but data remain variable depending on study technique, and at least one recent, large study demonstrated no decline in toxicity rates.
      • Haynes K
      • Heitjan DF
      • Kanetsky PA
      • Hennessy S.
      Declining public health burden of digoxin toxicity from 1991 to 2004.
      ,
      • Angraal S
      • Nuti SV
      • Masoudi FA
      • Freeman JV
      • Murugiah K
      • Shah ND
      • Desai NR
      • Ranasinghe I
      • Wang Y
      • Krumholz HM.
      Digoxin use and associated adverse events among older adults.
      ,
      • Hauptman PJ
      • Blume SW
      • Lewis EF
      • Ward S.
      Digoxin toxicity and use of digoxin immune Fab: insights From a National Hospital database.
      • Nuti SV
      • Masoudi FA
      • Freeman JV
      • Murugiah K
      • Desai NR
      • Ranasinghe I
      • Shah ND
      • Wang Y
      • Krumholz HM.
      Abstract 236: Trends in hospitalization for digoxin toxicity and subsequent outcomes among Medicare beneficiaries in the United States, 1999–2011.
      • See I
      • Shehab N
      • Kegler SR
      • Laskar SR
      • Budnitz DS.
      Emergency department visits and hospitalizations for digoxin toxicity: United States, 2005 to 2010.
      Toxicity typically occurs with digoxin concentrations >2 ng/ml, and adverse effects of toxicity are led by cardiac, neurologic, and gastrointestinal manifestations, leading to significant morbidity and mortality.
      • See I
      • Shehab N
      • Kegler SR
      • Laskar SR
      • Budnitz DS.
      Emergency department visits and hospitalizations for digoxin toxicity: United States, 2005 to 2010.
      ,
      • Yancy CW
      • Jessup M
      • Bozkurt B
      • Butler J
      • Casey DE
      • Drazner MH
      • Fonarow GC
      • Geraci SA
      • Horwich T
      • Januzzi JL
      • Johnson MR
      • Kasper EK
      • Levy WC
      • Masoudi FA
      • McBride PE
      • McMurray JJV
      • Mitchell JE
      • Peterson PN
      • Riegel B
      • Sam F
      • Stevenson LW
      • Tang WHW
      • Tsai EJ
      • Wilkoff BL
      American College of Cardiology Foundation, American Heart Association Task Force on Practice Guidelines. 2013 ACCF/AHA guideline for the management of heart failure: a report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines.
      Historically, all-cause mortality rates in patients with digoxin toxicity were as high as 20% to 30%, but more recently, in-hospital mortality has ranged from 3% to 20%.
      • Angraal S
      • Nuti SV
      • Masoudi FA
      • Freeman JV
      • Murugiah K
      • Shah ND
      • Desai NR
      • Ranasinghe I
      • Wang Y
      • Krumholz HM.
      Digoxin use and associated adverse events among older adults.
      ,
      • Hauptman PJ
      • Blume SW
      • Lewis EF
      • Ward S.
      Digoxin toxicity and use of digoxin immune Fab: insights From a National Hospital database.
      ,
      • Nuti SV
      • Masoudi FA
      • Freeman JV
      • Murugiah K
      • Desai NR
      • Ranasinghe I
      • Shah ND
      • Wang Y
      • Krumholz HM.
      Abstract 236: Trends in hospitalization for digoxin toxicity and subsequent outcomes among Medicare beneficiaries in the United States, 1999–2011.
      Digoxin immune fab (DIF) is an antidote therapy used in ∼20% cases of digoxin toxicity according to some estimates.
      • Hauptman PJ
      • Blume SW
      • Lewis EF
      • Ward S.
      Digoxin toxicity and use of digoxin immune Fab: insights From a National Hospital database.
      DIF therapy is well-tolerated with few side effects and has demonstrated strong pharmacological efficacy.
      • Antman EM
      • Wenger TL
      • Butler VP
      • Haber E
      • Smith TW.
      Treatment of 150 cases of life-threatening digitalis intoxication with digoxin-specific Fab antibody fragments. Final report of a multicenter study.
      • Wei S
      • Niu MT
      • Dores GM.
      Adverse events associated with use of digoxin immune fab reported to the US Food and Drug Administration Adverse Event Reporting System, 1986–2019.
      • Chan BSH
      • Buckley NA.
      Digoxin-specific antibody fragments in the treatment of digoxin toxicity.
      Less is known about the contemporary clinical outcomes of digoxin toxicity and DIF treatment. In this study, we aimed to better describe the characteristics, historic and contemporary incidence, and outcomes of patients presenting with suspected digoxin toxicity, either treated or not treated with DIF, in a large tertiary healthcare delivery system.

      Methods

      This study is a retrospective analysis of existing electronic health record (EHR) data collected from patients receiving medical care at Duke University Health System from 2000 to 2020. Incidence and temporal trends of suspected digoxin toxicity were analyzed. Suspected digoxin toxicity was defined as a hospital admission or emergency department visit with elevated digoxin concentrations (>2 ng/ml) and/or a primary diagnosis of digoxin toxicity (International Classification of Diseases Ninth/Tenth Revision, codes 972.1 or T46.0×1A) and/or an order for DIF. Administration of DIF and an order or prescription for DIF were recorded (including the medication names Digifab, Digibind, and DIF, and the Healthcare Common Procedure Coding System procedure code “J1162”). A representative denominator population was established by including all patients with a record of digoxin prescriptions each year to estimate suspected digoxin toxicity rates. Through a cohort study design, we then described the characteristics of this population and outcomes, stratified by treatment with DIF.
      The study population and the majority of variables were ascertained from Duke University Health System EHR data abstracted using the Duke Enterprise Data Unified Content Explorer (DEDUCE) query tool. DEDUCE is a web-based, clinical research query tool capable of searching various facets of the current and legacy electronic medical record systems. Mortality was ascertained using the death date recorded for the patient in DEDUCE. Echocardiogram results were sourced from the C3 database, which includes echocardiogram procedures conducted at Duke University Hospital and satellite clinics. Electrocardiogram results were sourced from data files maintained by the Duke Heart Center. A manual chart review of patients with full electronic medical records was performed to obtain additional details on clinical presentation and DIF treatment. Missing data were 0% for most of the multivariable model variables and <7% for all variables except heart rate, which was ascertained from the admission electrocardiogram and was missing in ∼40% of cases, due in part to data limitations in the early years of EHR availability. For digoxin concentration, estimated glomerular filtration rate (eGFR), and potassium, missing values were imputed with gender-specific medians. To address missing heart rate data, we categorized the variable and included “unknown” as one of the categories.
      The characteristics of patients with suspected digoxin toxicity were summarized at the encounter level, with potentially multiple hospital/emergency department encounters with suspected digoxin toxicity per patient. Factors associated with DIF treatment and clinical outcomes were assessed using a patient-level analysis, using the first eligible admission or emergency department visit during the study period. In-hospital 30-day and 1-year outcomes were assessed for patients receiving DIF therapy compared with matched controls with suspected digoxin toxicity. Each DIF-treated case was matched to 3 untreated controls. Matching was done using the logit of the estimated propensity to receive DIF. Variables were considered for inclusion in the propensity score model based on their likelihood to affect DIF use and outcomes, including the year of admission, age, gender, co-morbidities (HF, hypertension, AF), digoxin concentration, digoxin toxicity diagnosis, admission eGFR, admission potassium, and admission heart rate. Variables were selected for the final model using forward selection with p <0.10. The final multivariable propensity model included covariables for the year of admission, HF, digoxin serum concentration, admission potassium, and heart rate. Matching was done using an optimal matching algorithm, allowing for a maximum difference of 0.25 between the logit propensity score for cases versus matched controls. A total of 4 DIF-treated cases could not be matched and were excluded from the analysis.
      We used Fisher's exact test to compare binary outcomes (in-hospital mortality) and the Satterthwaite t test to compare the mean length of stay. We used the log-rank or Gray test (as appropriate) to compare the incidence of 30-day and 1-year outcomes between DIF-treated cases and those not treated with DIF. We compared the average rates of recurrent hospitalization events using robust Poisson regression. This research was approved by the Duke Health Institutional Review Board. All statistical analyses were performed using SAS, version 9.4 (SAS Institute, Cary, North Carolina). Two-sided p <0.05 were considered statistically significant.

      Results

      A total of 727 unique patients (779 total encounters) were identified as having suspected digoxin toxicity between 2000 and 2020. Of these patients, 93% had an elevated digoxin concentration (>2 ng/ml) and 3% had a primary diagnosis of digoxin toxicity by billing code (International Classification of Diseases Ninth/Tenth Revision, codes 972.1 or T46.0×1A). The remaining patients were included based on the administration of DIF and/or an order for DIF treatment. Digoxin concentrations varied widely but clustered in the 2 to 3 ng/ml range (Figure 1). DIF treatment was used in 67 cases (8.6%). Of the 25 cases with a primary diagnosis code of digoxin toxicity, 3 were treated with DIF (12.0%). A total of 713 cases had elevated digoxin concentrations but no diagnosis of toxicity billed, 24 of which were treated with DIF (3.4%). The majority of the DIF-treated cases (40 of 67 cases) were included in the cohort through an order or prescription for DIF and did not have a digoxin toxicity diagnosis billed or an elevated digoxin concentration.
      Figure 1
      Figure 1Distribution of digoxin concentrations for patients presenting with suspected or confirmed digoxin toxicity, stratified by treatment with DIF.
      A total of 12,432 unique patients with digoxin prescriptions were identified between 2009 and 2020 (a subset of the study period when outpatient prescription data were available). Digoxin prescriptions and the absolute frequency of suspected digoxin toxicity experienced a downtrend over time (Supplementary Table 1), whereas the incidence of digoxin toxicity per 1,000 patients with a digoxin prescription and the frequency of DIF treatment fluctuated without a clear trend (Figure 2).
      Figure 2
      Figure 2Trends in the incidence of suspected digoxin toxicity and the frequency of DIF treatment from 2009 to 2020. ED = emergency department.
      Primary diagnosis codes of admissions for patients treated with DIF were led by acute kidney failure and cardiac dysrhythmias, including AF. The diagnosis code entitled “Poisoning by cardiotonic/cardiac-stimulant glycosides” was noted as the primary diagnosis code in only 2 of the DIF-treated cases. Patients not treated with DIF were most commonly admitted with a primary diagnosis code for cardiac dysrhythmias, including AF, congestive HF, and acute kidney failure. Poisoning by cardiotonic/cardiac-stimulant glycosides was recorded as the primary diagnosis code for 26 patients in the non-treated group.
      The baseline characteristics of patients with suspected digoxin toxicity, stratified by DIF treatment, are shown in Table 1. Patients with suspected digoxin toxicity were mostly White (72%) and women (62%) and had an average age of 72 ± 16 years. These patients had a high burden of co-morbidities, namely hypertension (53%), AF (48%), and HF (44%). The most common concomitant cardiovascular medication was a β blocker (24%). Digoxin dose was ∼250 µg/day in 43% of the cases and ∼125 µg/day in 34% of the total cases. More patients were admitted to the hospital (93%) compared with patients discharged from the emergency department (7%) (Table 2).
      Table 1Baseline patient characteristics
      All cases

      (N = 779)
      Cases with DIF treatment

      (N = 67)
      Cases without DIF treatment

      (N = 712)
      P-value
      Demographics & Comorbidities
      Age (years)74.5 (64.5, 82.8)77.1 (67.8, 83.6)74.5 (64.4, 82.7)0.28
      Women479 (62%)40 (60%)439 (62%)0.75
      White550 (72%)46 (71%)504 (72%)
      Black195 (26%)16 (25%)179 (26%)
      Other19 (3%)3 (5%)16 (2%)
      Hypertension408 (52%)44 (66%)364 (51%)0.023
      Diabetes247 (32%)22 (33%)225 (32%)0.84
      Myocardial infarction117 (15%)13 (19%)104 (15%)0.29
      Stroke/TIA45 (6%)4 (6%)41 (6%)1.00
      Peripheral vascular disease137 (18%)16 (24%)121 (17%)0.16
      Renal disease176 (23%)15 (22%)161 (23%)0.97
      Hyperlipidemia205 (26%)21 (31%)184 (26%)0.33
      Chronic pulmonary disease245 (32%)20 (30%)225 (32%)0.77
      Atrial fibrillation373 (48%)39 (58%)334 (47%)0.077
      Atrial flutter52 (7%)4 (6%)48 (7%)1.00
      Heart failure343 (44%)39 (58%)304 (43%)0.014
      Admission vital signs
      Body mass index (kg/m2)25.5 (22.2, 30.4)25.6 (23.1, 30.5)25.5 (22.2, 30.4)0.88
      Heart rate
      Heart rate from clinical reads and supplemented with ECG.
      (bpm)
      76.0 (63.0, 91.0)53.0 (39.5, 69.5)77.0 (64.0, 91.5)<.001
      Systolic BP (mmHg)111.5 (98.0, 135.5)126.0 (98.0, 142.0)110.0 (98.0, 130.0)0.37
      Diastolic BP (mmHg)65.0 (58.5, 73.0)58.0 (50.0, 63.0)66.0 (59.0, 74.0)0.002
      ECG findings
      ECG data available492 (63.2%)35 (52.2%)457 (64.2%)0.053
      Concern for digoxin effect15/492 (3.0%)1/35 (2.9%)14/457 (3.1%)1.00
      Atrial fibrillation110/492 (22.4%)7/35 (20.0%)103/457 (22.5%)0.73
      Atrial flutter26/492 (5.3%)1/35 (2.9%)25/457 (5.5%)1.00
      Ectopic atrial rhythm7/492 (1.4%)0/35 (0.0%)7/457 (1.5%)1.00
      Sinus bradycardia26/492 (5.3%)5/35 (14.3%)21/457 (4.6%)0.030
      First-degree AV block35/492 (7.1%)6/35 (17.1%)29/457 (6.3%)0.030
      Second-degree AV block60/492 (12.2%)8/35 (22.9%)52/457 (11.4%)0.058
      Third-degree AV block1/492 (0.2%)1/35 (2.9%)0/457 (0.0%)0.071
      Junctional rhythm13/492 (2.6%)2/35 (5.7%)11/457 (2.4%)0.24
      Non-specific AV block9/492 (1.8%)1/35 (2.9%)8/457 (1.8%)0.49
      PVC67/492 (13.6%)6/35 (17.1%)61/457 (13.3%)0.61
      Ventricular tachycardia1/492 (0.2%)0/35 (0.0%)1/457 (0.2%)1.00
      Digoxin exposure
      Digoxin daily dose on/prior to admission (mcg/day)0.65
       <12517/138 (12.3%)2/16 (12.5%)15/122 (12.3%)
       ∼12547/138 (34.1%)8/16 (50.0%)39/122 (32.0%)
       ∼25059/138 (42.8%)6/16 (37.5%)53/122 (43.4%)
      High loading doses9/138 (6.5%)0/16 (0.0%)9/122 (7.4%)
      One-time injections6/138 (4.3%)0/16 (0.0%)6/122 (4.9%)
      Highest serum digoxin concentration during admission (ng/mL)2.5 (2.2, 3.0)2.7 (1.7, 3.8)2.5 (2.2, 3.0)0.83
      Admission cardiovascular medications
      Beta blocker138/566 (24%)11/46 (24%)127/520 (24%)0.94
      ACEi/ARB86/566 (15%)11/46 (24%)75/520 (14%)0.086
      MRA27/566 (5%)0/46 (0%)27/520 (5%)0.16
      Calcium channel blocker69/566 (12%)10/46 (22%)59/520 (11%)0.039
      Statin87/566 (15%)6/46 (13%)81/520 (16%)0.65
      Loop diuretics120/566 (21%)15/46 (33%)105/520 (20%)0.048
      Admission laboratory measures
      Serum creatinine (mg/dL)1.4 (1.1, 2.2)1.8 (1.2, 3.2)1.4 (1.0, 2.1)0.014
      eGFR,
      eGFR calculated from creatinine using the CKD-EPI equation (2009) without race modifier. ACEi/ARB = angiotensin-converting enzyme inhibitor/angiotensin receptor blocker; AV = atrioventricular; BP = blood pressure; CKD-EPI = Chronic Kidney Disease Epidemiology Collaboration; ECG = electrocardiogram; eGFR = estimated glomerular filtration rate; LVEF = left ventricular ejection fraction; MRA = magnetic resonance angiography; PVC = premature ventricular contraction; TIA = transient ischemic attack.
      calculated without race modifier (ml/min/1.73 m2)
      39.1 (23.6, 60.9)30.3 (14.8, 48.6)40.0 (24.2, 61.2)0.013
      Serum potassium (mEq/L)4.3 (3.9, 4.8)4.5 (4.0, 5.3)4.3 (3.9, 4.8)0.022
      Echocardiographic measures
      Echo data available285 (36.6%)20 (29.9%)265 (37.2%)0.23
      LVEF %)0.052
       <3067/285 (23.5%)1/20 (5%)66/265 (25%)
       30-4047/285 (16.5%)3/20 (15%)44/265 (17%)
       >40-<5539/285 (13.7%)6/20 (30%)33/265 (13%)
       >=55132/285 (46.3%)10/20 (50%)122/265 (46%)
      Values are the median (25th, 75th percentiles) or n (%).
      low asterisk Heart rate from clinical reads and supplemented with ECG.
      eGFR calculated from creatinine using the CKD-EPI equation (2009) without race modifier.ACEi/ARB = angiotensin-converting enzyme inhibitor/angiotensin receptor blocker; AV = atrioventricular; BP = blood pressure; CKD-EPI = Chronic Kidney Disease Epidemiology Collaboration; ECG = electrocardiogram; eGFR = estimated glomerular filtration rate; LVEF = left ventricular ejection fraction; MRA = magnetic resonance angiography; PVC = premature ventricular contraction; TIA = transient ischemic attack.
      Table 2Hospitalization characteristics and discharge laboratory measures
      All cases

      (N = 779)
      Cases with DIF treatment

      (N = 67)
      Cases without DIF treatment

      (N = 712)
      P-value
      Characteristics of admission
      ED visit (versus hospital admission)53 (6.8%)0 (0.0%)53 (7.4%)0.011
      Duration of time in ED,
      For both ED visits and hospitalizations.
      hours
      5.8 (4.3, 8.2)5.7 (4.6, 7.5)5.9 (4.3, 8.3)0.858
      Discharge laboratory measures (for patients discharged alive)
      Serum creatinine, mg/dL1.2 (1.0, 1.7)1.2 (1.0, 1.4)1.3 (1.0, 1.8)0.601
      eGFR
      eGFR calculated from creatinine using the CKD-EPI equation (2009) without race modifier. BNP = brain natriuretic peptide; CKD-EPI = Chronic Kidney Disease Epidemiology Collaboration; ED = emergency department; eGFR = estimated glomerular filtration rate; NT-proBNP = N-terminal pro-brain natriuretic peptide.
      48.0 (31.8, 65.9)47.7 (37.9, 66.3)48.1 (31.3, 65.8)0.597
      Serum potassium, mEq/L4.2 (3.8, 4.5)4.2 (4.0, 4.5)4.2 (3.8, 4.5)0.424
      Values are the median (25th, 75th percentiles) or n (%).
      low asterisk For both ED visits and hospitalizations.
      eGFR calculated from creatinine using the CKD-EPI equation (2009) without race modifier.BNP = brain natriuretic peptide; CKD-EPI = Chronic Kidney Disease Epidemiology Collaboration; ED = emergency department; eGFR = estimated glomerular filtration rate; NT-proBNP = N-terminal pro-brain natriuretic peptide.
      Patients treated with DIF demonstrated a higher burden of co-morbidities, particularly hypertension (Table 1). In addition, these patients had significantly lower presenting heart rates (median 53 vs 77 beats/minute, p <0.001), worse renal function (median eGFR, 30.3 vs 40.0 ml/min/1.73 m2, p = 0.013), and higher potassium (median 4.5 vs 4.3 mEq/L, p = 0.022). However, most patients with elevated potassium were not treated with DIF therapy; 17% (n = 6) of patients with admission potassium >5.5 mEq/L (and digoxin concentrations >2 ng/ml) and 18%, n = 3, of those with admission potassium >6.0 mEq/L (and digoxin concentrations >2 ng/ml) were treated with DIF (Supplementary Table 2). In addition to higher rates of sinus bradycardia, patients treated with DIF had trends toward more frequent atrioventricular block (Table 1). Patients treated with DIF experienced a similar duration of care in the emergency department (median 5.7 vs 5.9 hours, p = 0.858). At discharge, renal function and potassium in patients treated with DIF had improved on average to become equivalent to that in the non-treated group (Table 2).
      In the univariable analysis, the following factors were noted to have a significant association with DIF treatment: a history of HF, hypertension, AF, digoxin serum concentration, admission potassium, and admission heart rate (Supplementary Table 3). In the multivariable model, year of admission, a history of HF, digoxin concentration (primarily 2.2 to 4 ng/ml vs ≥4 ng/ml), admission potassium (>5.2 mEq/L vs 3.6 to 5.2 mEq/L), and heart rate (primarily <60 beats/min vs 60 to 100 beats/min, and “unknown” vs 60 to 100 beats/min) emerged as significant predictors of DIF treatment (Table 3). The overall C-statistic of the multivariable model was 0.78 (95% confidence interval 0.72 to 0.84).
      Table 3Factors associated with DIF treatment in patients presenting with suspected digoxin toxicity by multivariable analysis
      DescriptionDFWald chi-squarePR > chi-squareOdds ratio (OR) estimateLower 95% confidence limit for ORUpper 95% confidence limit for OR
      Year of admission27.610.022
       2007-2013 vs 2000-20060.4630.2470.868
       2014-2020 vs 2000-20061.3380.5773.102
      Heart failure16.600.010
       Yes vs no2.0731.1893.615
      Digoxin serum concentration29.040.011
       <2.2 ng/mL vs ≥4 ng/mL0.5170.1821.469
       2.2-<4 ng/mL vs ≥4 ng/mL0.2790.1130.690
      Admission potassium27.820.020
       <3.6 mEq/L vs 3.6-5.2 mEq/L0.3700.0851.609
       >5.2 mEq/L vs 3.6-5.2 mEq/L2.3801.1494.932
      Admission ECG HR326.26<.001
       <60 bpm vs 60-100 bpm6.2972.82414.044
       >100 bpm vs 60-100 bpm0.2490.0302.051
       Unknown vs 60-100 bpm3.6471.7447.627
      DF = degrees of freedom; ECG = electrocardiogram; HR = heart rate; PR = probability.
      In the complete cohort of 727 patients, 92 patients (12.7%) died during their index hospitalization: 7.7% in the DIF therapy cohort and 13.1% in the cohort without DIF treatment (p = 0.245; Table 4). The mean and median lengths of stay (including patients discharged alive and those deceased in the hospital) were 14.7 and 8 days, respectively. Starting from admission, the cumulative mortality incidence was 15.3% at 30 days (n = 111 deaths) and 43% at 1 year (n = 304 deaths). The rate of all-cause hospitalization was high, with 21.4% readmitted within 30 days of discharge and 48.9% readmitted within 1 year. The event rate, counting recurrent hospital admissions within 1 year of discharge, was 141.5 admissions per 100 patient-years. Overall, 64.7% of patients had experienced all-cause readmission or death by 1-year postdischarge.
      Table 4Outcomes of patients treated with DIF compared with patients without DIF treatment (first admission or ED visit per patient)
      Before matchingAfter matching
      With DIF treatment

      (N = 65)
      Without DIF treatment

      (N = 662)
      P-valueWith DIF treatment

      (N = 61)
      Matched controls

      (N = 183)
      P-value
      In-hospital outcomes
      In-hospital death5/65 (7.7%)87/662 (13.1%)0.2455/61 (8.2%)29/183 (15.8%)0.199
      Length of stay, N6561461173
       mean (std)11.7 (15.8)15.0 (21.3)0.12211.6 (16.3)13.9 (21.3)0.392
      Mortality on or after admission
      Mortality
       30-day incidence8 (12.3%)103 (15.6%)0.5037 (11.5%)36 (19.7%)0.153
       1-year incidence26 (40.8%)278 (42.9%)0.63225 (41.9%)84 (46.7%)0.415
       Event rate,
      For death and composite endpoints, the event rate was calculated for the first event only. Each DIF-treated case was matched to 3 untreated controls. Matching was done using the logit of the estimated propensity to receive DIF. The multivariable propensity model included covariates for year of admission, heart failure, digoxin serum concentration, admission potassium, and heart rate. Matching was done using an optimal matching algorithm (%dist and %vmatch SAS macros from Mayo Clinic), allowing for a maximum difference of 0.25 between the logit propensity score for cases versus matched controls. Four DIF-treated cases could not be matched and were excluded from this analysis. CI = confidence interval; DIF = digoxin immune Fab; HF = heart failure.
      per 100 patient-years (95% CI)
      57.5 (37.9 - 87.5)63.9 (55.9 - 72.9)0.64359.5 (38.9 - 91.1)71.6 (56.0 - 91.4)0.462
      1-year outcomes post-discharge
      among patients discharged alive, N6057556154
      Mortality
       30-day incidence5 (8.3%)63 (11.0%)0.5324 (7.1%)16 (10.4%)0.482
       1-year incidence21 (35.9%)191 (34.3%)0.86120 (36.7%)55 (36.7%)0.988
       Event rate,
      For death and composite endpoints, the event rate was calculated for the first event only. Each DIF-treated case was matched to 3 untreated controls. Matching was done using the logit of the estimated propensity to receive DIF. The multivariable propensity model included covariates for year of admission, heart failure, digoxin serum concentration, admission potassium, and heart rate. Matching was done using an optimal matching algorithm (%dist and %vmatch SAS macros from Mayo Clinic), allowing for a maximum difference of 0.25 between the logit propensity score for cases versus matched controls. Four DIF-treated cases could not be matched and were excluded from this analysis. CI = confidence interval; DIF = digoxin immune Fab; HF = heart failure.
      per 100 patient-years (95% CI)
      48.7 (30.9 - 76.9)46.6 (40.0 - 54.3)0.85549.9 (31.3 - 79.5)49.6 (37.5 - 65.8)0.984
      All-cause hospitalization
       30-day incidence8 (13.3%)128 (22.3%)0.1118 (14.3%)38 (24.7%)0.112
       1-year incidence32 (54.0%)270 (48.4%)0.64930 (54.1%)85 (56.5%)0.615
       Event rate, per 100 patient-years (95% CI)153.1 (110.0 - 213.1)140.2 (123.2 - 159.7)0.628149.8 (106.5 - 210.5)191.4 (152.1 - 240.8)0.242
      HF hospitalization
       30-day incidence0 (0.0%)29 (5.0%)0.2210 (0.0%)12 (7.8%)0.868
       1-year incidence12 (20.2%)77 (13.7%)0.19311 (19.9%)30 (19.9%)0.973
       Event rate, per 100 patient-years (95% CI)34.8 (19.6 - 61.7)32.9 (25.6 - 42.3)0.86332.5 (17.7 - 59.5)54.2 (36.0 - 81.5)0.170
      All-cause hospitalization/mortality
       30-day incidence11 (18.3%)173 (30.1%)0.06310 (17.9%)47 (30.5%)0.078
       1-year incidence41 (69.2%)360 (64.3%)0.81138 (68.5%)105 (69.5%)0.664
       Event rate,
      For death and composite endpoints, the event rate was calculated for the first event only. Each DIF-treated case was matched to 3 untreated controls. Matching was done using the logit of the estimated propensity to receive DIF. The multivariable propensity model included covariates for year of admission, heart failure, digoxin serum concentration, admission potassium, and heart rate. Matching was done using an optimal matching algorithm (%dist and %vmatch SAS macros from Mayo Clinic), allowing for a maximum difference of 0.25 between the logit propensity score for cases versus matched controls. Four DIF-treated cases could not be matched and were excluded from this analysis. CI = confidence interval; DIF = digoxin immune Fab; HF = heart failure.
      per 100 patient-years (95% CI)
      95.1 (74.0 - 122.3)87.8 (80.4 - 95.9)0.55694.9 (73.0 - 123.2)94.8 (81.1 - 110.7)0.996
      HF hospitalization/mortality
       30-day incidence5 (8.3%)88 (15.3%)0.6364 (7.1%)25 (16.2%)0.651
       1-year incidence28 (47.6%)232 (41.4%)0.46026 (47.5%)72 (47.8%)0.909
       Event rate,
      For death and composite endpoints, the event rate was calculated for the first event only. Each DIF-treated case was matched to 3 untreated controls. Matching was done using the logit of the estimated propensity to receive DIF. The multivariable propensity model included covariates for year of admission, heart failure, digoxin serum concentration, admission potassium, and heart rate. Matching was done using an optimal matching algorithm (%dist and %vmatch SAS macros from Mayo Clinic), allowing for a maximum difference of 0.25 between the logit propensity score for cases versus matched controls. Four DIF-treated cases could not be matched and were excluded from this analysis. CI = confidence interval; DIF = digoxin immune Fab; HF = heart failure.
      per 100 patient-years (95% CI)
      65.0 (45.1 - 93.6)56.6 (49.6 - 64.6)0.48764.9 (44.4 - 94.9)65.0 (51.7 - 81.7)0.995
      low asterisk For death and composite endpoints, the event rate was calculated for the first event only.Each DIF-treated case was matched to 3 untreated controls. Matching was done using the logit of the estimated propensity to receive DIF. The multivariable propensity model included covariates for year of admission, heart failure, digoxin serum concentration, admission potassium, and heart rate. Matching was done using an optimal matching algorithm (%dist and %vmatch SAS macros from Mayo Clinic), allowing for a maximum difference of 0.25 between the logit propensity score for cases versus matched controls. Four DIF-treated cases could not be matched and were excluded from this analysis.CI = confidence interval; DIF = digoxin immune Fab; HF = heart failure.
      In the unadjusted analysis, no differences were observed in the outcomes (mortality/hospitalization) between patients treated with DIF and those not treated with DIF (Table 4). The matching algorithm resulted in 2 well-balanced cohorts (Supplementary Table 4). In the matched multivariable analysis, patients treated with DIF experienced numerically lower mean length of stay (11.6 vs 13.9 days, p = 0.39), numerically lower rates of adjusted, in-hospital mortality (8.2% vs 15.8%, p = 0.20), and numerically lower rates of all-cause rehospitalization (14.3% vs 24.7%, p = 0.11) compared with a matched cohort (Figure 3, Table 4). The 2 groups demonstrated similar postdischarge hospitalization and mortality outcomes 1 year postdischarge (Figure 3, Table 4).
      Figure 3
      Figure 3Mortality and all-cause hospitalization outcomes at 1 year by patients treated with DIF versus matched controls. (A) 1-year postadmission cumulative incidence of death by DIF treatment. Population: patients treated with DIF versus matched controls (first encounter only). (B) 1-year postdischarge cumulative incidence of all-cause hospitalization by DIF treatment. Population: patients treated with DIF versus matched controls (first encounter only). CI = confidence interval.

      Discussion

      This study demonstrates several key findings regarding digoxin utilization, suspected digoxin toxicity, and DIF therapy. First, although digoxin prescriptions and absolute digoxin toxicity rates have decreased over time,
      • Haynes K
      • Heitjan DF
      • Kanetsky PA
      • Hennessy S.
      Declining public health burden of digoxin toxicity from 1991 to 2004.
      ,
      • Angraal S
      • Nuti SV
      • Masoudi FA
      • Freeman JV
      • Murugiah K
      • Shah ND
      • Desai NR
      • Ranasinghe I
      • Wang Y
      • Krumholz HM.
      Digoxin use and associated adverse events among older adults.
      ,
      • Hauptman PJ
      • Blume SW
      • Lewis EF
      • Ward S.
      Digoxin toxicity and use of digoxin immune Fab: insights From a National Hospital database.
      ,
      • Nuti SV
      • Masoudi FA
      • Freeman JV
      • Murugiah K
      • Desai NR
      • Ranasinghe I
      • Shah ND
      • Wang Y
      • Krumholz HM.
      Abstract 236: Trends in hospitalization for digoxin toxicity and subsequent outcomes among Medicare beneficiaries in the United States, 1999–2011.
      the incidence per prescription has fluctuated, as would be expected with a relatively rare event, without any clear downtrend. This is consistent with much of the existing literature and extends our understanding of incidence per prescription, which has been a challenge to capture in studies to date and may not be decreasing as significantly as absolute toxicity rates. DIF use was relatively modest at <10% of cases with fluctuation but without any clear trends up or down over the past 20 years. DIF was used in patients with a higher co-morbidity burden (namely, a history of HF) and a more acutely ill presentation, representing a potentially sicker population. Patients treated with DIF were typically discharged with normal renal function and potassium, although this analysis can only present this observation as opposed to comment on any causative relationship. Patients treated with DIF also demonstrated nonsignificant, numerically lower in-hospital mortality and 30-day all-cause rehospitalization. Notably, all 30-day hospitalization/mortality event rates were lower in the DIF-treated group; a larger cohort may be able to further assess the certainty and effect size of this in-hospital trend. Of note, no evidence of an increased HF risk was found in the setting of DIF use (a theoretical concern of “removing” the digoxin HF effect) as evidenced by no increase in length of stay and a numerically lower HF hospitalization risk postdischarge. Outcomes between DIF-treated patients and matched controls were equivalent by 1-year postdischarge.
      The role of digoxin in clinical practice has notably changed over time and, accordingly, trends and perceptions of digoxin toxicity have evolved. The randomized Digitalis Investigation Group trial produced neutral results in 1997, and digoxin has been associated with increased mortality in several meta-analyses.
      • Vamos M
      • Erath JW
      • Hohnloser SH.
      Digoxin-associated mortality: a systematic review and meta-analysis of the literature.
      ,
      • Vamos M
      • Erath JW
      • Benz AP
      • Lopes RD
      • Hohnloser SH.
      Meta-analysis of effects of digoxin on survival in patients with atrial fibrillation or heart failure: an update.
      These results along with the drug's narrow therapeutic window, the risk of adverse side effects, and competing pharmacologic options have led to a decrease in use over time.
      • Ambrosy AP
      • Bhatt AS
      • Stebbins AL
      • Wruck LM
      • Fudim M
      • Greene SJ
      • Kraus WE
      • O'Connor CMO
      • Piña IL
      • Whellan DJ
      • Mentz RJ
      Prevalent digoxin use and subsequent risk of death or hospitalization in ambulatory heart failure patients with a reduced ejection fraction - findings from the Heart Failure: a Controlled Trial Investigating Outcomes of Exercise Training (HF-ACTION) randomized controlled trial.
      It has been noted, though, that these meta-analyses were limited by bias within the studies.
      • Kotecha D.
      Digoxin vs bisoprolol for heart rate control in atrial fibrillation-reply.
      ,
      • Ziff OJ
      • Lane DA
      • Samra M
      • Griffith M
      • Kirchhof P
      • Lip GYH
      • Steeds RP
      • Townend J
      • Kotecha D.
      Safety and efficacy of digoxin: systematic review and meta-analysis of observational and controlled trial data.
      Furthermore, there is mixed evidence in other recent large secondary analyses regarding the association of digoxin with mortality.
      • Ambrosy AP
      • Bhatt AS
      • Stebbins AL
      • Wruck LM
      • Fudim M
      • Greene SJ
      • Kraus WE
      • O'Connor CMO
      • Piña IL
      • Whellan DJ
      • Mentz RJ
      Prevalent digoxin use and subsequent risk of death or hospitalization in ambulatory heart failure patients with a reduced ejection fraction - findings from the Heart Failure: a Controlled Trial Investigating Outcomes of Exercise Training (HF-ACTION) randomized controlled trial.
      The recent RATE-AF (Rate Control Therapy Evaluation in Permanent Atrial Fibrillation) trial, which compared β blocker and digoxin therapy in patients with permanent AF with HF symptoms, has also reinvigorated interest in the potential of digoxin.
      • Kotecha D
      • Bunting KV
      • Gill SK
      • Mehta S
      • Stanbury M
      • Jones JC
      • Haynes S
      • Calvert MJ
      • Deeks JJ
      • Steeds RP
      • Strauss VY
      • Rahimi K
      • Camm AJ
      • Griffith M
      • Lip GYH
      • Townend JN
      • Kirchhof P.
      Rate Control Therapy Evaluation in Permanent Atrial Fibrillation (RATE-AF) Team. Effect of digoxin vs bisoprolol for heart rate control in atrial fibrillation on patient-reported quality of life: the RATE-AF randomized clinical trial.
      Overall, digoxin use may remain modest, but it seems that the drug might continue to serve a select population with HF and/or AF at this time.
      The persistent utility of digoxin in select patients will drive the ongoing significance of digoxin toxicity. Modulation of toxicity risk and prevention of toxicity remain the optimal approach. This includes careful consideration of necessity and dosing in patients who are older, women, have reduced renal function, and/or have a low body mass index.
      • Yancy CW
      • Jessup M
      • Bozkurt B
      • Butler J
      • Casey DE
      • Drazner MH
      • Fonarow GC
      • Geraci SA
      • Horwich T
      • Januzzi JL
      • Johnson MR
      • Kasper EK
      • Levy WC
      • Masoudi FA
      • McBride PE
      • McMurray JJV
      • Mitchell JE
      • Peterson PN
      • Riegel B
      • Sam F
      • Stevenson LW
      • Tang WHW
      • Tsai EJ
      • Wilkoff BL
      American College of Cardiology Foundation, American Heart Association Task Force on Practice Guidelines. 2013 ACCF/AHA guideline for the management of heart failure: a report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines.
      ,
      • Jelliffe RW
      • Brooker G.
      A nomogram for digoxin therapy.
      Our study findings support these practice pattern considerations as patients with toxicity were generally older (median age 74.5 [64.5 to 82.8] years), mostly women (61.5%), and often on “higher-dose” digoxin (i.e., 43% on ∼250 µg/day and 6.5% with high loading doses).
      In the setting of digoxin toxicity, DIF therapy has been established as being safe and efficacious in terms of treatment response (i.e., resolution of signs and symptoms of toxicity, improvement in free, active digoxin and potassium concentrations) with low adverse event rates.
      • Antman EM
      • Wenger TL
      • Butler VP
      • Haber E
      • Smith TW.
      Treatment of 150 cases of life-threatening digitalis intoxication with digoxin-specific Fab antibody fragments. Final report of a multicenter study.
      ,
      • Wei S
      • Niu MT
      • Dores GM.
      Adverse events associated with use of digoxin immune fab reported to the US Food and Drug Administration Adverse Event Reporting System, 1986–2019.
      DIF utilization overall is fairly uncommon but is associated with an estimated 8,000 hospital visits annually in the United States.
      • Hauptman PJ
      • Blume SW
      • Lewis EF
      • Ward S.
      Digoxin toxicity and use of digoxin immune Fab: insights From a National Hospital database.
      Data are conflicting as to whether DIF therapy is increasing with some recent studies demonstrating stable utilization at ∼21% to 22%.
      • Hauptman PJ
      • Blume SW
      • Lewis EF
      • Ward S.
      Digoxin toxicity and use of digoxin immune Fab: insights From a National Hospital database.
      ,
      • Hussain Z
      • Swindle J
      • Hauptman PJ.
      Digoxin use and digoxin toxicity in the post-DIG trial era.
      Overall, our study suggests that DIF is used in a minority of cases with potential for digoxin toxicity (∼9% in our study, although determination of suspected digoxin toxicity differs by study) without clear trends toward an increase or decrease in use over time.
      This study carries the intrinsic limitations of a single-center, retrospective analysis. Importantly, this analysis assessed suspected digoxin toxicity in the total cohort, and some of these patients may have not met the strict criteria for signs and symptoms of traditionally defined digoxin toxicity. We included the total cohort of patients with digoxin concentrations >2 ng/ml and/or a primary diagnosis of digoxin toxicity and/or a DIF order for 2 reasons: (1) our interest in assessing a cohort of possible digoxin toxicity with the potential for consideration of DIF therapy, and (2) the inability to fully assess all subtle signs and symptoms of digoxin toxicity in any retrospectively designed study.
      In addition, our study period extends back to the early days of EHRs; during this time, lab and medication data may have been more incomplete than in the contemporary era. Missing data had the potential to affect the following variables: digoxin concentration (4.4% missing), eGFR (6.5%), potassium (6.7%), and heart rate (39.6%). The former three were handled by imputation using gender-specific medians and heart rate by including an “unknown” category. Also of note, a manual chart review of recent patients with available EHR charts revealed that some patients had a DIF prescription ordered but not given (i.e., the order was placed and then canceled); this review resulted in 2 of 12 patients being re-classified to the non-DIF group. It is possible that additional historical patients (in whom a manual chart review was not possible) also had DIF therapy prescribed but not given.
      ?A3B2 tlsb=-0.01w?>In summary, digoxin toxicity remains a pertinent public health issue with implications for morbidity and mortality despite a reduction in digoxin utilization. Continued efforts in mitigating risk of digoxin toxicity with selective and low-dose utilization is important to target a reduction in risk per prescription. DIF therapy appears to be typically used in a medically complex population with a higher acuity of illness at presentation and is associated with nonsignificant, numerically lower in-hospital mortality and 30-day all-cause rehospitalization compared with a matched cohort not treated with DIF therapy. These trends dissipate over time and result in similar outcomes by 1-year postdischarge. In addition, there was no evidence of increased HF risk in the setting of DIF use as supported by no increase in length of stay and numerically lower HF hospitalization risk postdischarge.

      Disclosures

      Dr. Ambrosy has received relevant research support through grants to his institution from Amarin Pharma, Inc., Abbott, and Novartis; and has received modest reimbursement for travel from Novartis. Dr. Fudim receives consulting fees from Abbott, Audicor, AxonTherapies, Bodyguide, Bodyport, Boston Scientific, CVRx, Daxor, Edwards Lifesciences, Feldschuh Foundation, Fire1, Gradient, Intershunt, NXT Biomedical, Pharmacosmos, PreHealth, Shifamed, Splendo, Vironix, Viscardia, and Zoll. The remaining authors have no conflicts of interest to declare.

      Appendix. Supplementary materials

      References

        • Haynes K
        • Heitjan DF
        • Kanetsky PA
        • Hennessy S.
        Declining public health burden of digoxin toxicity from 1991 to 2004.
        Clin Pharmacol Ther. 2008; 84: 90-94
        • Ambrosy AP
        • Bhatt AS
        • Stebbins AL
        • Wruck LM
        • Fudim M
        • Greene SJ
        • Kraus WE
        • O'Connor CMO
        • Piña IL
        • Whellan DJ
        • Mentz RJ
        Prevalent digoxin use and subsequent risk of death or hospitalization in ambulatory heart failure patients with a reduced ejection fraction - findings from the Heart Failure: a Controlled Trial Investigating Outcomes of Exercise Training (HF-ACTION) randomized controlled trial.
        Am Heart J. 2018; 199: 97-104
        • Angraal S
        • Nuti SV
        • Masoudi FA
        • Freeman JV
        • Murugiah K
        • Shah ND
        • Desai NR
        • Ranasinghe I
        • Wang Y
        • Krumholz HM.
        Digoxin use and associated adverse events among older adults.
        Am J Med. 2019; 132: 1191-1198
        • Kotecha D
        • Bunting KV
        • Gill SK
        • Mehta S
        • Stanbury M
        • Jones JC
        • Haynes S
        • Calvert MJ
        • Deeks JJ
        • Steeds RP
        • Strauss VY
        • Rahimi K
        • Camm AJ
        • Griffith M
        • Lip GYH
        • Townend JN
        • Kirchhof P.
        Rate Control Therapy Evaluation in Permanent Atrial Fibrillation (RATE-AF) Team. Effect of digoxin vs bisoprolol for heart rate control in atrial fibrillation on patient-reported quality of life: the RATE-AF randomized clinical trial.
        JAMA. 2020; 324: 2497-2508
        • Hauptman PJ
        • Blume SW
        • Lewis EF
        • Ward S.
        Digoxin toxicity and use of digoxin immune Fab: insights From a National Hospital database.
        JACC Heart Fail. 2016; 4: 357-364
        • Nuti SV
        • Masoudi FA
        • Freeman JV
        • Murugiah K
        • Desai NR
        • Ranasinghe I
        • Shah ND
        • Wang Y
        • Krumholz HM.
        Abstract 236: Trends in hospitalization for digoxin toxicity and subsequent outcomes among Medicare beneficiaries in the United States, 1999–2011.
        Circ Cardiovasc Qual Outcomes. 2015; 8: A236
        • See I
        • Shehab N
        • Kegler SR
        • Laskar SR
        • Budnitz DS.
        Emergency department visits and hospitalizations for digoxin toxicity: United States, 2005 to 2010.
        Circ Heart Fail. 2014; 7: 28-34
        • Yancy CW
        • Jessup M
        • Bozkurt B
        • Butler J
        • Casey DE
        • Drazner MH
        • Fonarow GC
        • Geraci SA
        • Horwich T
        • Januzzi JL
        • Johnson MR
        • Kasper EK
        • Levy WC
        • Masoudi FA
        • McBride PE
        • McMurray JJV
        • Mitchell JE
        • Peterson PN
        • Riegel B
        • Sam F
        • Stevenson LW
        • Tang WHW
        • Tsai EJ
        • Wilkoff BL
        American College of Cardiology Foundation, American Heart Association Task Force on Practice Guidelines. 2013 ACCF/AHA guideline for the management of heart failure: a report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines.
        J Am Coll Cardiol. 2013; 62: e147-e239
        • Antman EM
        • Wenger TL
        • Butler VP
        • Haber E
        • Smith TW.
        Treatment of 150 cases of life-threatening digitalis intoxication with digoxin-specific Fab antibody fragments. Final report of a multicenter study.
        Circulation. 1990; 81: 1744-1752
        • Wei S
        • Niu MT
        • Dores GM.
        Adverse events associated with use of digoxin immune fab reported to the US Food and Drug Administration Adverse Event Reporting System, 1986–2019.
        Drugs Real World Outcomes. 2021; 8: 253-262
        • Chan BSH
        • Buckley NA.
        Digoxin-specific antibody fragments in the treatment of digoxin toxicity.
        Clin Toxicol (Phila). 2014; 52: 824-836
        • Vamos M
        • Erath JW
        • Hohnloser SH.
        Digoxin-associated mortality: a systematic review and meta-analysis of the literature.
        Eur Heart J. 2015; 36: 1831-1838
        • Vamos M
        • Erath JW
        • Benz AP
        • Lopes RD
        • Hohnloser SH.
        Meta-analysis of effects of digoxin on survival in patients with atrial fibrillation or heart failure: an update.
        Am J Cardiol. 2019; 123: 69-74
        • Kotecha D.
        Digoxin vs bisoprolol for heart rate control in atrial fibrillation-reply.
        JAMA. 2021; 325: 1681-1682
        • Ziff OJ
        • Lane DA
        • Samra M
        • Griffith M
        • Kirchhof P
        • Lip GYH
        • Steeds RP
        • Townend J
        • Kotecha D.
        Safety and efficacy of digoxin: systematic review and meta-analysis of observational and controlled trial data.
        BMJ. 2015; 351: h4451
        • Jelliffe RW
        • Brooker G.
        A nomogram for digoxin therapy.
        Am J Med. 1974; 57: 63-68
        • Hussain Z
        • Swindle J
        • Hauptman PJ.
        Digoxin use and digoxin toxicity in the post-DIG trial era.
        J Card Fail. 2006; 12: 343-346