Literature Search

To obtain an evidence basis for the analysis, a literature search was performed to identify all randomized controlled trials (RCTs; level 1 evidence) dealing with treatment of mild diabetic foot infections. Two researchers (E.V.L., P.M.) independently performed a comprehensive search of the MEDLINE, EMBASE, and Cochrane databases using the key terms “diabetes or diabetic” and “foot or lower limb” and “ulcer or infection or cellulitis”. Studies of both IV and oral antibiotics versus an active comparator were included. As well, references from all retrieved articles and reviews were searched by hand. Discrepancies were settled through consensus; if the 2 researchers could not achieve consensus, a third reviewer was enlisted to make the final decision. Pharmacoeconomic Model
The information from RCTs and the 2004 IDSA guide­lines was used to construct a decision tree (Figure 1) to determine which antibiotic regimens were cost-effective in treating mild diabetic foot infections. TreeAge Pro 2007 software (TreeAge Software Inc, Williamstown, Massachusetts) was used for this purpose. As well, 3 Canadian infectious diseases experts, based in Toronto, Ontario, and Vancouver, British Columbia, were consulted (personal communications, October 29, 2007); these experts confirmed that they followed the IDSA guidelines for the treatment of diabetic foot infections.

The population for the model consisted of patients with type 1 or type 2 diabetes and acute but mild foot infections who were able to take oral antibiotics. Mild infections were defined as those limited to the skin or superficial tissues with the presence of 2 or more manifestations of inflammation (e.g., purulence, erythema, pain, tenderness, warmth, or induration) or any cellulitis up to 2 cm around the ulcer; patients with systemic illnesses, osteomyelitis, gangrene, or extensive deep tissue infections were excluded. Conversely, if oral antibiotic therapy failed, the patient’s infection was considered moderate to severe. Moderate to severe infections were defined as cellulitis extending beyond 2 cm, deep tissue involvement, gangrene, and/or involvement of muscle, joint, tendon, or bone requiring admission to hospital for debridement and parenteral administration of antibiotics.
canada viagra online

The first branch of the decision tree (Figure 1) reflects the oral antibiotics recommended by the IDSA as empiric therapy for the treatment of mild diabetic foot infections. Sulfamethoxazole-trimethoprim was excluded from the model because no RCTs involving this drug were found. The remaining antibiotics—cephalexin, clindamycin, cloxacillin (in place of dicloxacillin, which is no longer available in Canada), amoxicillin-clavulanate, and levofloxacin—were included. The regimens for these antibiotics, as reported in the RCTs and the IDSA guidelines, are listed in Table 1.

Figure 1. Decision tree for treatment

Figure 1. Decision tree for treatment of mild diabetic foot infections in Canada. Definitions of symbols: +, expandable branch; square, decision node; circle, chance node; triangle, terminal node; #, calculated probability value based on other model inputs.

Probabilities for clinical success were obtained from the RCTs identified. However, only the probabilities of clinical success for evaluable patients were used in the decision tree, because probability values from the literature were most complete for this group. Clinical success was defined as both microbiological and clinical resolution of the infection. Clinical success rates for amoxicillin-clavulanate were not avail­able; therefore, the probability was extrapolated from cure rates. If more than one RCT studied the drug of interest, the average of the reported clinical success rates was used as the probability in the model. A list of event probabilities used in the decision- tree model is presented in Table 2.
canadian discount pharmacy

Table 1. Model Inputs: Doses and Durations of Antibiotic Therapy for Mild Diabetic Foot Infections


Antibiotic


Dose


Duration (days)


Reference


Oral

Amoxicil lin-clavu la nate

500
mg q8h

10

Lipsky
and others

Cephalexin

500
mg qid

10

Lipsky
and others

Clindamycin

300
mg qid

10

Lipsky
and others

Cloxacillin

1000
mg qid

10

Lipsky
and others

Levofloxacin

750
mg daily

10

Graham
and others


Parenteral

Imipenem

500
mg q6h

14

Grayson
and others

Bouter
and others

Piperacillin-tazobactam

3.375
g q6h

14

Lipsky
and others

Ticarcillin-clavulanate

3.1 g q6h

14

Graham
and others

Tan and others

The next branch of the decision tree reflects the pathways of either clinical success or clinical failure after a 10-day course of oral antibiotics. If clinical success occurs, then treatment ends. If clinical failure occurs, the patient is admitted to hospital for IV administration of antibiotics (on the basis of expert opinion). Antibiotics for the treatment of moderate to infections suggested by the infectious disease experts included imipenem, piperacillin-tazobactam, ticarcillin- clavulanate, ciprofloxacin (or another fluoroquinolone) plus clindamycin or metronidazole, and ceftriaxone (or another third-generation cephalosporin) plus clindamycin or metronidazole. Ertapenem was also mentioned by the experts, but only for finishing a course of treatment on an outpatient basis, because of the convenience of once-daily dosing, provided the microorganisms involved are susceptible. From the list of antibiotics provided by the infectious diseases experts, only the antibiotic regimens that had been evaluated in RCTs were included in the model: imipenem, piperacillin-tazobactam, and ticarcillin-clavulanate. Treatment duration was 14 days, as suggested in the IDSA guidelines provided there was no bone involvement. Three outcomes were possible after 14 days of IV antibiotic therapy: clinical success, amputation, or death. The probabilities of clinical success for the secondary treatments (i.e., following failure of primary drug therapy) were also obtained from RCTs, and the probability of death was obtained from an abstract and a recent systematic review of dia­betic foot infections. Probabilities of amputation were calcu­lated from the probabilities of clinical success and death within that branch of the decision-tree model. That is, if patients were not cured and did not die, it was assumed that amputation would be necessary. The total time horizon of the model was 24 days: 10 days for the primary antibiotic therapy and an addi­tional 14 days for subsequent secondary antibiotic therapy, if the primary therapy failed.
viagra 50 mg

Table 2. Inputs: Clinical Success Rates and Costs


Variables


No. of Study


Probability


Cost


Reference


Participants


(95% CI)*


(2007 Can$)


Clinical success


Amoxicillin-clavulanate


108


0.7130 (0.621-0.790)


Lipsky and others


Cephalexin


29


0.8621 (0.694-0.945)


Lipsky and others


Clindamycin


27


0.9630 (0.817-0.993)


Lipsky and others


Cloxacillin


27


0.7037 (0.515-0.841)


Lipsky and others


Levofloxacin


26


0.6923 (0.500-0.835)


Graham and others


Imipenem


69


0.9033 (0.788-0.940)


Grayson
and others


Bouter
and others


Piperacillin-tazobactam


196


0.8265 (0.767-0.873)


Lipsky and others


Ticarcillin—clavulanate


37


0.7857 (0.515-0.804)


Graham
and others


Tan and
others


Infection leading to death


183


0.0098


Nelson and others,


Amato and
others


Cost


Amoxicillin-clavulanate

$0.67/tablet (500 mg amoxicillin, 125 mg clavulanate)


ODB/CDI


Cephalexin


$0.24/500-mg tablet


ODB/CDI


Clindamycin


$0.78/300-mg capsule


ODB/CDI


Cloxacillin


$0.19/500-mg capsule


ODB/CDI


Levofloxacin


$5.19/500-mg tablet


ODB/CDI


Imipenem


$26.00/500-mg vial


McKesson Canada


Piperacillin-tazobactam


$18.00/3.375-g vial


McKesson Canada


Ticarcillin-clavulanate



$11.00/3.1

-g vial


McKesson Canada


Lower extremity amputation


$12,334.37/


CIHI CMGs


(except toe)


procedure


Pronouncement of death



$31
.45/death


MOHLTC


Admission to hospital, medical


bed


$430.30/day


Sunnybrook


Admission to hospital, surgical


bed


$517.05/day


Sunnybrook


Admission to hospital, ICU bed



$2041
.67/day


Sunnybrook


CI = confidence interval; CIHI CMGs = Canadian Institute for Health Information Case Mix Groups; ICU = intensive
care unit;

MOHLTC = Ontario
Ministry of Health and Long-Term Care; ODB/CDI


=

Ontario Drug Benefit Formulary / Comparative Drug Index.


All costs are reported in 2007 Canadian dollars. Only direct costs were included in this model, to reflect the perspec­tive of the Ontario Ministry of Health and Long-Term Care. Drug costs were determined from the 2007 Ontario Drug Ben­efit Formulary/Comparative Drug Index and McKesson Canada. Amputation costs were obtained from the Canadian Institute for Health Information database using case mix group (CMG) methodology for the fiscal year 2002/2003; values were adjusted to 2007 by means of the Canadian Consumer Price Index. The CMG methodology is designed to aggregate acute inpatient data in terms of various Canadian resources, such as the procedural costs in the model created in this study. Only the direct costs of lower-extremity amputations (excluding the toe) were used in the model. Physician fees were obtained from the 2007 Ontario Schedule of Benefits for Physician Services. Hospital costs were obtained from the 2003/2004 database of the Sunnybrook Health Sciences Centre27 and were adjusted to 2007 with the Consumer Price Index. Costs of admission to a medical, surgical, or intensive care unit were obtained, but only the cost of admission to a medical bed (excluding overhead costs) was used in calculating the costs of 14 days of parenteral antibiotic therapy in hospital. Costs were not discounted, as the duration of all treatment periods was less than 1 year. Table 2 summarizes the costs of antibiotics, hospitalization, amputation, and death that were input into the model.
silagra 100

Category: Diabetes / Tags: antibiotic, Canada, cost-effectiveness, diabetic foot infection, oral therapy, pharmacoeconomics

Leave a Reply

Your email address will not be published. Required fields are marked *

CAPTCHA image
*