	Table of contents

	Current issue

	Search journal

	Archived issues

	NZMJ Obituaries 1887-2008

	Classifieds

	Hotline (free ads)

	How to subscribe

	How to contribute

	How to advertise

	Contact Us

	Copyright

	Other journals

Journal of the New Zealand Medical Association, 12-December-2003, Vol 116 No 1187

Digoxin therapeutic drug monitoring: an audit and review

Andrew Sidwell, Murray Barclay, Evan Begg and Grant Moore

Abstract

Aim The measurement and assessment of digoxin concentrations are often performed poorly. We have conducted an audit to assess the appropriateness of digoxin therapeutic drug monitoring in Christchurch Hospital.

Methods One hundred consecutive requests for digoxin concentrations in Christchurch Hospital inpatients were assessed. The case notes and hospital medication records were reviewed to determine the indication for testing, the appropriateness of the sampling time and of the subsequent alteration to dosing.

Results In 53% of requests no clear indication for digoxin therapeutic drug monitoring (TDM) could be determined. In the remainder, ‘suspected toxicity’ accounted for 31% and ‘therapeutic failure’ for 16%. Samples were inappropriately taken within eight hours post-dose in 32% of requests. In 19% of cases, the samples did not reflect steady-state conditions. In 5% of occasions, the subsequent decision regarding dose adjustment was felt to be clearly inappropriate, and there was uncertainty regarding appropriateness in some other cases. Overall, in only 29% of requests was TDM performed appropriately with regard to indication, sampling and subsequent dose alteration.

Conclusions At Christchurch Hospital, the practice of TDM for digoxin is often inappropriate. It would seem that medical staff education is required to improve this practice.

Digoxin is currently indicated largely for control of ventricular rate in atrial fibrillation, but it is also used as a positive inotropic agent in congestive cardiac failure. An assay for measuring serum digoxin concentration became available in 1969, and therapeutic drug monitoring (TDM) for digoxin followed. The upper limit of the recommended therapeutic range (1.0 to 2.5 nmol/l) reflects the significant increase in risk of toxicity that occurs with serum concentrations over 2.6 nmol/l. Above 3.8 nmol/l, toxicity is almost invariable.1 Prior to the availability and implementation of digoxin TDM, the reported incidence of digoxin toxicity in all patients receiving digoxin therapy ranged from 8 to 29%.2 The introduction of TDM to aid decisions regarding digoxin therapy has been associated with a significant reduction in toxicity.3

The indications for digoxin TDM are relatively few and include confirmation of clinically suspected toxicity, assessing the reasons for therapeutic failure, assessing medication compliance, and assessing the effects of factors that alter the pharmacokinetics of digoxin (predominantly renal dysfunction and drug interactions). Samples for digoxin TDM are required to be taken at least eight hours post-dose to allow for the redistribution of digoxin from plasma into the tissues and thus avoid erroneous interpretation of elevated results that may mislead management decisions. In certain indications (therapeutic failure and assessing pharmacokinetic influences) it is ideal that digoxin concentrations represent a steady-state situation. The relatively long half-life of digoxin (30 hours minimum) means that following dose alterations a steady-state situation takes at least one week to achieve (often significantly longer in the elderly). Studies have demonstrated that digoxin TDM is frequently performed inappropriately,4–6 but there are no data for New Zealand. An audit was conducted on requests for digoxin concentrations to assess the appropriateness of indication, sample timing and subsequent clinical response to digoxin TDM in Christchurch Hospital.

Methods

One hundred consecutive digoxin TDM requests to Canterbury Health Laboratories in September and October 2002 were assessed for information on digoxin dose, dose time and sample time. The requests for digoxin TDM were identified by laboratory staff and forwarded to the audit team. Only requests on inpatients were assessed due to potential difficulty obtaining information on outpatients. The patients’ hospital case notes and medication records were then examined by the clinical pharmacology registrar to determine the indication for testing, the duration of digoxin therapy, the appropriateness of the sample time and the appropriateness of the subsequent alteration to dose. The criteria of appropriateness were established prior to the audit by the authors, from the Clinical Pharmacology and Toxicology departments. The indication for TDM was determined from the clinical notes, either directly from written comments or indirectly from comments suggesting symptoms of toxicity or therapeutic failure. The indication was classified as appropriate if there was possible clinical toxicity, therapeutic failure, or a possibility of changing digoxin concentrations due to changing renal function or a drug interaction. The blood sampling time was classified as inappropriate if within eight hours post-dose, or before steady-state had been reached. Dose adjustments were classified as inappropriate if they were likely to result in either a toxic or subtherapeutic digoxin concentration, or if there was inappropriate continuation or cessation of therapy. Clinically significant results were discussed with the managing medical teams by the clinical pharmacology registrar with advice provided as per the usual manner for TDM within Christchurch Hospital. Ethics approval was not sought as it was felt that this study represented audit only.

Results

The 100 requests for digoxin TDM were performed in 75 patients (1 request in 60 patients, 2 requests in 8 patients, 3 requests in 4 patients, and 4 requests in 3 patients). The results regarding the appropriateness of the indication and sample timing are shown in Table 1.

Table 1. Appropriateness of digoxin TDM indication and sample timing (n = 100)

	Appropriate (%)	Inappropriate (%)
Indication Toxicity Therapeutic failure None	31 16	53
Sample timing Post-distribution (>8 h) Steady-state	68 81	32 19

An appropriate indication for testing was present in 47% (31% for confirmation of toxicity and 16% for therapeutic failure). In the remaining 53% no clear indication was found. The dose and dose time were provided on the request form in only 12% and 14% of requests respectively, whereas the sample time was provided in 84% of requests. Appropriate post-distribution samples (more than eight hours following a digoxin dose) were taken in 68% of requests, with the remaining 32% of samples taken too early (within eight hours). Steady-state in serum concentrations (greater than four half-lives of digoxin after dose initiation or change) was not reached in 19%.

The reported therapeutic range for digoxin at Canterbury Health Laboratories and in most laboratories is 1.0 to 2.5nmol/l. Digoxin concentrations were low in 34% of requests, ‘therapeutic’ in 53%, and high in 13%. The relationship between TDM indication and measured digoxin concentration in these patients is shown in Table 2.

Table 2. Relationship between TDM indication and digoxin concentration (n = 100)

Digoxin concentration	TDM indication
Digoxin concentration	Toxicity (%)	Therapeutic failure (%)	None (%)
Low Normal High	39 42 19	31 63 6	32 57 11

As a result of TDM, there was no change to the digoxin regimen in 67% of cases, a dose increase in 5%, dose reduction in 10%, and discontinuation of digoxin therapy in 18% of requests. The relationship between the digoxin concentration and the subsequent alteration to the digoxin regimen is shown in Table 3.

Table 3. Relationship between digoxin concentration and subsequent dose alteration (n = 100)

Digoxin concentration	Alteration to digoxin dose
Digoxin concentration	Discontinued (%)	Reduced (%)	None (%)	Increased (%)
Low Normal High	18 9 54	0 8 46	71 81 0	12 2 0

The alteration to the digoxin regimen was felt to be clearly inappropriate in 5% of requests. The inappropriate alterations included one case in which the dose was reduced despite a low therapeutic digoxin concentration. On three occasions the dose was reduced following the measurement of high digoxin concentrations, when samples were taken within eight hours of the preceding digoxin dose and there were no symptoms of toxicity in the patients. A further case of inappropriate dose alteration occurred when digoxin had been withheld appropriately following toxicity, but the digoxin remained withheld despite a subsequent digoxin concentration measurement in the lower end of the therapeutic range.

Seven alterations to digoxin dose following the return of a high concentration could possibly be considered inappropriate. These requests were in patients for whom there was no evidence of toxicity (the indication was either therapeutic failure or not identified), but digoxin was discontinued in two, and the dose reduced in the other five in response to the result. For patients with a low digoxin concentration, 71% had no alteration to their digoxin dose. It was difficult to determine in these patients whether this decision was inappropriate or not. Overall, in only 29% of requests was TDM performed correctly in terms of appropriate indication, sampling time and subsequent dose alteration. However, even in these correctly performed requests, none provided all the desired information (dose, dose time and sample time) on the request form.

Discussion

Although therapeutic drug monitoring of digoxin has been practised for more than 30 years, it appears that it is often performed inappropriately. One potential explanation is that the majority of requests are made by inexperienced junior medical staff, although this explanation was not assessed specifically in this study. Over half of the requests did not appear to have an appropriate indication, other than appearing to be a routinely ordered along with other investigations.

The clinical suspicion of toxicity correlated poorly with high digoxin concentrations. In those requests in which the indication for TDM was confirmation of toxicity, only 19% were associated with a high digoxin concentration. This result is not necessarily surprising given that many of the symptoms of digoxin toxicity (nausea, confusion, arrhythmias, and abdominal pain) are non-specific and are frequently present in acutely ill patients in general.

Serum digoxin concentrations should be interpreted within the clinical context. It was apparent in our study that no change in dose occurred in most patients with low concentrations. The appropriateness of this is difficult to determine, but it is likely that on a number of occasions dose adjustment should have occurred. First, the need for digoxin should be questioned. If the indication for therapy is rate control and the current ventricular rate is appropriate in the presence of low digoxin concentrations then a trial without digoxin may be appropriate. If rate control is unsatisfactory, this may be related to other acute illness processes, and treatment of the underlying condition may be all that is required. Alternatively, it may be that the dose of digoxin needs to be increased. The validity of the therapeutic range in determining efficacy is unclear. In terms of improving rate control in chronic atrial fibrillation, only a weak correlation between digoxin concentration and ventricular rate was found in a review of the literature.7 This is not surprising given the number of other influences on the atrio-ventricular node, such as altered sympathetic drive with other comorbidities (eg, sepsis, hypoxia). However, TDM for individual patients may be useful to detect the patients who have a low digoxin concentration and who may benefit from an increase in digoxin dose, as opposed to those with higher concentrations who are likely to develop toxicity symptoms only from an increase in dose.

In congestive cardiac failure there is increasing evidence that concentrations lower than the currently recommended limit of the therapeutic range (<1.0 nmol/l) may be as efficacious or even better than higher concentrations. Adams et al recently demonstrated that serum concentrations of digoxin did not correlate with clinical outcome.8 In a combined analysis of the PROVED and RADIANCE trials, they found that clinical outcome for heart failure was the same if the digoxin concentration was low (0.6 to 1.2 nmol/l), medium (1.2 to 1.5 nmol/l) or high (>1.5 nmol/l). The results for all three groups were superior to placebo. There were a number of methodological limitations in the PROVED and RADIANCE trials. Patients were not randomly assigned to the three serum digoxin concentration groups, the sample size was small and follow up was for two months only. The multivariate analysis did not include renal function, which is likely to affect the digoxin concentrations and may also reflect prognosis of heart failure. Despite these criticisms, lower serum digoxin concentrations (0.6 to 1.2 nmol/l) may be as efficacious as and less pro-arrhythmic than higher concentrations in patients with heart failure.

A recent post hoc analysis of the DIG trial takes this further, suggesting that lower serum digoxin concentrations are associated with better mortality and hospitalisation outcomes compared with higher concentrations in men with heart failure and sinus rhythm.9 Lower concentrations (0.6 to 1.0 nmol/l) were associated with a small but significant reduction in all-cause mortality, worsening heart failure, all-cause hospitalisation and hospitalisation due to heart failure compared with placebo. Clinical outcome with mid-range concentrations (1.1 to 1.4 nmol/l) was not significantly different from placebo, and higher concentrations (>1.5 nmol/l) were associated with a small but significant increase in all-cause mortality, cardiovascular mortality and hospitalisation for digoxin-related toxicity. The number of women in the original DIG trial was too low to allow for adequate power analysis to replicate the findings in men. However, there was a trend towards the same results. This may suggest that the results of another previous post hoc analysis of the DIG trial,10 which suggested an increase in mortality rate for women treated with digoxin for heart failure, may be dependent upon serum concentration. Ideally, further research is required to define more comprehensively the relationship between efficacy and serum concentration, for both men and women.

In this study, high digoxin concentrations were generally deemed by the prescribers to represent toxicity, and the digoxin dose was either reduced or discontinued regardless of whether symptoms of toxicity were present or not. An elevated concentration in itself does not necessarily imply toxicity. It could, however, be argued that because the benefits of a higher concentration are questionable, the dose should be reduced anyway. On the other hand, it is important to recognise that toxicity is not excluded by concentrations within the therapeutic range. Abnormal plasma potassium and calcium concentrations, and hyper- or hypothyroidism affect tissue sensitivity to digoxin and alter the therapeutic index.

In approximately one third of patients, samples were collected within eight hours of a digoxin dose, and therefore provided inaccurate results. One of the difficulties at Christchurch Hospital is that the four daily phlebotomy rounds all occur in the eight hours following the most common dose time (0800h) for patients receiving digoxin. In another report, a hospital policy of changing the routine dose time of digoxin from 0900h to 1300h resulted in only 2/118 samples being taken within eight hours of a digoxin dose.11 However, this represented only a small improvement in this particular hospital, as there were few sampling errors prior to the intervention. The majority of patients on digoxin are elderly and compliance may be affected by changes to routine. Such changes are therefore not recommended. It would seem more appropriate to improve the practice of health professionals regarding TDM use, and to arrange sampling at relevant times, such as early morning prior to medication rounds.

In summary, although the cost of performing serum digoxin concentrations is relatively small in terms of the entire laboratory budget, appropriate requisitions can contribute to cost savings. At the present time it is clear that, in Christchurch Hospital, requests are frequently performed inappropriately. All therapeutic drug monitoring has limitations in assisting clinical decision making. An incorrect decision in any step of the process of TDM can result in an inappropriate clinical action. It is important for clinicians to be aware of the correct indications and methods for digoxin concentration monitoring, and to act appropriately on the results.

Author information: Andrew I Sidwell, Clinical Pharmacology Registrar; Murray L Barclay, Clinical Pharmacologist; Evan J Begg, Clinical Pharmacologist, Department of Clinical Pharmacology, Christchurch Hospital; Grant A Moore, Head of Department of Toxicology, Canterbury Health Laboratories, Christchurch Hospital, Christchurch

Correspondence: Dr Murray Barclay, Department of Clinical Pharmacology, Christchurch Hospital, Private Bag 4710, Christchurch. Fax: (03) 364 1003; email: MurrayB@cdhb.govt.nz

References:

Smith TW, Haber E. Digoxin Intoxication: the relationship of clinical presentation to serum digoxin concentration. J Clin Invest 1970;49:2377–86.
Beller GA, Smith TW, Abelmann WH, et al. Digitalis intoxication. A prospective clinical study with serum level correlations. N Engl J Med 1971;284:989–97.
Duhme DW, Greenblatt DJ, Koch-Weser J. Reduction of digoxin toxicity associated with measurement of serum levels. A report from the Boston Collaborative Drug Surveillance Program. Ann Intern Med 1974;80:516–9.
Williamson KM, Thrasher KA, Fulton KB, et al. Digoxin toxicity: an evaluation in current clinical practice. Arch Intern Med 1998;158:2444–9.
Makela EH, Davis SK, Piveral K, et al. Effect of data collection method on results of serum digoxin concentration audit. Am J Hosp Pharm 1988;45:126–30.
Mayan H, Bloom E, Hoffman A. Use of digoxin monitoring in a hospital setting as an essential tool in optimizing therapy. J Pharm Technol 2002;18:133–6.
Masuhara JE, Lalonde RL. Serum digoxin concentrations in atrial fibrillation: a review. Drug Intell Clin Pharm 1982;16:543–6.
Adams KF Jr, Gheorghiade M, Uretsky BF, et al. Clinical benefits of low serum digoxin concentrations in heart failure. J Am Coll Cardiol 2002;39:946–53.
Rathore SS, Curtis JP, Wang Y, et al. Association of serum digoxin concentration andoutcomes in patients with heart failure. JAMA 2003;289:871–8.
Rathore SS, Wang Y, Krumholz HM. Sex-based differences in the effect of digoxin for the treatment of heart failure. N Engl J Med 2002;347:1403–11.
Fraser GL, Wennberg DE, Dickens JD Jr, Lambrew CT. Changing physician behaviour in ordering digoxin assays. Ann Pharmacother 1996;30:449–54.