TMP-SMZ: Be Careful

1 Comment

altYou rarely hear of problems with the old standby, TMP-SMZ, but there is a growing body of literature suggesting that physicians need to be wary of the drug’s potentially serious side effects.

You rarely hear of problems with the old standby, TMP-SMZ, but there is a growing body of literature suggesting that physicians need to be wary of the drug’s potentially serious side effects.

Trimethoprim – sulfamethoxazole (hereinafter, TMP-SMZ – and also known as co-trimazole by the Brits) is undoubtedly among the most commonly prescribed antibiotics in the ED setting. In the past, its use was largely limited to the treatment of bladder infections. But now, in the age of MRSA infections causing most skin abscesses, TMP-SMZ has risen to one of the key antibiotics to treat these infections.

Generally TMP-SMZ is considered a very safe drug. You rarely hear of allergies to it or other problems. But not so fast. There is a growing literature that suggests that those prescribing this drug need an increased level of respect for the potential serious side effects caused by this combination of drugs.

Although there are some new drugs for anticoagulating the estimated 2.3+ million people who have atrial fibrillation, it is likely that it will take a long time to convert patients to them and, as such, emergency physicians will still be seeing patients taking warfarin for years to come.

Why the concern? Because antibiotics predictably cause patients on warfarin to have increases in their INR levels – and some are more capable of doing this than others. The risk for major hemorrhage on warfarin is about 2% per year and the case fatality rate has been estimated at 13%. All emergency physicians have seen the serious bleeding complications associated with warfarin. What we don’t want to do is to be the cause of hemorrhage in our patients on warfarin because of uninformed prescribing

TMP-SMZ seems to be among the worst of the culprit antibiotics in that it not only effects bacteria in the gut that are needed for the production of vitamin K (required for the hepatic synthesis of clotting factors II, VII, IX and X), but it competes with warfarin for catabolism via the famous P450 pathway.

Here’s a study that dramatically demonstrates the effects of TMP-SMZ compared with some other antibiotics (most antibiotics can effect gut flora and effect vitamin K production). In short, within days of initiation of antibiotic treatment the mean increase in INR with azithromycin was 0.5 with 31% of patients being increased to outside the therapeutic range; levofloxacin use raised the INR by a mean of 0.8 pushing 39% outside the therapeutic range and TMP-SMZ increased INR levels by 1.8 pushing 69% outside the therapeutic range with 31% having an INR of 5 or higher and 13% having bleeding complications. Although the 13% figure sounds very high, the point is that TMP-SMZs effects are substantial and consistent.

Glasheen, J.J., et al, J Gen Intern Med 20(7):653, July 2005

BACKGROUND: Although an interaction between warfarin and various antibiotics has been cited as a cause of overanticoagulation, there are only limited data on specific changes in the INR associated with antibiotic use.

METHODS: The authors, from the Denver VA Medical Center, retrospectively evaluated changes in the INR in patients with a stable INR while on a maintenance warfarin regimen, who were then prescribed terazosin (a 29-patient control group, as terazosin has no documented interaction with warfarin), azithromycin (32 patients), levofloxacin (27 patients) or trimethoprim-sulfamethoxazole (TMP/SMX) (16 patients).

RESULTS: The median duration of follow-up was 6-7 days in the antibiotic groups and 17 days in the terazosin group, and the mean patient age was 70. There was a mean INR decrease of 0.1 in the terazosin group, while the INR increased by a mean of 0.5 in the azithromycin group, 0.8 in the levofloxacin group and 1.8 in the TMP/SMX group. Elevation of the INR above the therapeutic range was noted in 31% and 39% of the azithromycin and levofloxacin groups, and in 69% of the TMP/SMX group. Corresponding rates of INRs above 4 were 16%, 19% and 44%, respectively. In the TMP/SMX group, 38% exhibited a mean INR increase of 2 or more points, 31% had an INR of 5 or higher, and 13% developed an adverse bleeding event (compared with none of the patients in the other groups).

CONCLUSIONS: This small study quantitates changes in the INR associated with the use of azithromycin, levofloxacin and (especially) TMP/SMX among patients treated with warfarin. 18 references 11/05 – #27

The following study indicates that lowering the warfarin dose preemptively can limit the risk of overanticoagulation with prescription of TMP-SMZ. But look what happened to the INR when there was no reduction in the warfarin dose – the INR was above 4 in 8 of 9 patients and above 6 in 4 of 9.

Ahmed, A., et al, J Thromb Thrombolysis 26(1):44, August 2008

BACKGROUND: Because some antibiotics can potentiate the effects of warfarin, patients on maintenance warfarin anticoagulation often undergo repeat INR testing within several days after initiation of these agents. Some experts favor preemptive reduction in the warfarin dose when patients are treated with these antibiotics.

METHODS: In this controlled study, from the University of Michigan, 40 patients on maintenance warfarin prophylaxis who required treatment with trimethoprim-sulfamethoxazole (TMP-SMX) or levofloxacin (antibiotics known to potentiate warfarin) were assigned to be managed with preemptive warfarin dose reduction by 10-20% (18 patients, chosen at the discretion of an anticoagulation clinic nurse) or to usual care with early measurement of the INR after starting the course of antibiotics (22 patients).

RESULTS: The mean baseline INR was about 2.5 in both groups. The mean warfarin dose reduction in the intervention group was about 16%. The INR was rechecked after a mean of 5.1 days in the intervention group and 4.7 days in controls. Among patients treated with TMP-SMX, the follow-up INR was above 4.0 in 2/8 intervention group patients (25%) but in 8/9 controls (90%), and it was above 6.0 in 4/9 control patients. Among patients treated with levofloxacin, the follow-up INR was above 4.0 in 0/10 intervention group patients but in 5/13 controls (39%); the INR was subtherapeutic in 4/10 intervention group patients (40%) taking levofloxacin.

CONCLUSIONS: The authors suggest that preemptive warfarin dose reduction appears to be an effective strategy in patients requiring a course of TMP-SMX, but that follow-up INR testing after about five days might be more appropriate in those started on a course of levofloxacin. 15 references ( – no reprints) 1/09 – #22

Want some other reasons to be careful with TMP-SMZ? In a wide ranging study of drug-drug interactions (DRUG-DRUG INTERACTIONS AMONG ELDERLY PATIENTS HOSPITALIZED FOR DRUG TOXICITY Juurlink, D.N., et al, JAMA 289(13):1652, April 2, 2003), it was noted that
among 179,986 patients taking glyburide, the adjusted odds ratio (OR) for hospital admission for hypoglycemia within one, two or three weeks after exposure to co-trimoxazole was 6.6, 5.7 and 4.9, respectively, when compared with glyburide-treated control patients who did not develop hypoglycemia. So hypoglycemia requiring hospitalization is another concern when prescribing TMP-SMZ. Other studies have shown similar findings with glipizide and, in fact, a wide range of antibiotics have been associated with hypoglycemia with the theory being that the infection that is being treated is somehow related.

TMP-SMZ is also at the top of the antibiotic list for causing Steven-Johnson syndrome and toxic epidermal necrolysis.  Specifically the crude relative risk with TMP-SMZ has been reported as 172 (with a confidence interval of 75 to 396)

Hyperkalemia is a common concern in the ED. Here again, TMP-SMZ can be a significant contributor. The following paper noted a nearly seven fold increase in admissions for hyperkalemia in elderly patients on ACE inhibitors or ARBs who were concomitantly prescribed TMP-SMZ. There are multiple case reports of TMP-SMZ also causing severe hyponatremia.

Antoniou, T., et al, Arch Intern Med 170(12):1045, June 28, 2010

BACKGROUND: Both ACE inhibitors and angiotensin receptor blockers (ARBs) are associated with a risk for development of hyperkalemia. This risk can be exacerbated by the co-administration of drugs that impair potassium excretion, such as trimethoprim-sulfamethoxazole (TMP-SMX), which is structurally similar to the potassium-sparing diuretic amiloride and reduces urinary potassium excretion by about 40%.

METHODS: The authors, from St. Mary’s Hospital in Toronto, performed a population-based case-control study to evaluate the relationship between treatment with an ACE inhibitor or ARB and development of hyperkalemia in patients aged 66 or older who were prescribed an antibiotic commonly used for the treatment of urinary tract infection (TMP-SMX, amoxicillin, norfloxacin, ciprofloxacin or nitrofurantoin). The study included 367 case patients who were hospitalized for hyperkalemia within 14 days after beginning antibiotic treatment and 1,417 matched control patients also receiving an ACE inhibitor or ARB in addition to one of the study antibiotics but who were not hospitalized for hyperkalemia.

RESULTS: When compared with treatment with amoxicillin as the reference antibiotic, the adjusted odds ratio (OR) for hospital admission due to hyperkalemia in association with treatment with TMP-SMX was 6.7 (95% CI 4.5-10.0). In contrast, ORs for a hyperkalemia-related hospital admission with use of one of the other antibiotics was 0.8-1.1 (with all 95% CIs crossing 1.0). CONCLUSIONS: This study demonstrates a nearly seven-fold increase in the risk of hospitalization for hyperkalemia in elderly patients managed with an ACE inhibitor or ARB who are prescribed trimethoprim-sulfamethoxazole. As such, it would be prudent to avoid the use of TMP-SMX in this population when other antibiotic options are available. 23 references ( – no reprints) 11/10 – #22

Some patients are particularly prone to adverse effects produced by TMP-SMZ. AIDS patients being treated for Pneumocystis carinii with TMP-SMZ are at substantial increased risk of skin reactions when on this combination drug with approximately 40% demonstrating rashes.

TMP-SMZ can also cause thrombocytopenia, leukopenia and neutropenia. Often times febrile children put on TMP-SMZ will have elevated white counts initially but on repeat testing the following day or so have more normalized white counts. It could easily be assumed that the reduction in white count is cause by the resolving of the infection by the antibiotic while it is quite likely that the reduction in the count is due to the hematologic effects of TMP-SMZ.   

Bottom line – particular care should be taken when prescribing TMP-SMZ to patients on warfarin, diabetics on sulfonylureas, elderly patients on ACEIs and ARBs (a growing trend in the treatment of congestive failure, post-myocardial infarctions and hypertension) and HIV patients.

Richard Bukata, MD Editor of Emergency Medical Abstracts (

1 Comment

Leave A Reply