Home / Resources / Articles / Cost Effectiveness of Preventing Diabetic Foot Ulcers

Cost Effectiveness of Preventing Diabetic Foot Ulcers

Apr 1, 2017

Does an ounce of prevention beat a pound of cure?

According to the ADA, treatment of diabetic foot ulcers (DFUs) along with associated infections, below the knee amputations, and surgeries to revascularize the lower limbs account for a significant portion of the costs incurred in the treatment of diabetes. Yet with the frequency of occurrence of these complications, there are very few studies that drive the paradigm toward either primary prevention (avoiding DFUs entirely) or secondary/tertiary measures (efficient treatment of DFUs in those who are not aware [secondary]/are aware [tertiary] of diabetic ulcers), which are combined into a single term (secondary prevention) for purposes of the article. Sadly, utilization of primary prevention of these complications is spotty in most health care systems, and implementation of secondary prevention is often delayed in patients with DFUs. It is speculated that one reason little attention is paid to these secondary measures may be the concern over a “small return on the investment” in trying to prevent amputations, an attitude that certainly appears to be both counter-intuitive and counterproductive. An attempt to show otherwise was made by N.R. Barshes et al. who utilized a Markov model demonstrating the probability of significant cost savings attributable to otherwise less costly preventive measures.


The idea of the Markov model allows prediction of transition from one condition to another, with the understanding that the probability of any transition is only dependent on the current condition, but not any past condition, and that these conditions exist over a continuum. A simple example would be the states of untreated, treated, and final outcomes (cure, amputation, or death, the latter two of which would be considered “inescapable” outcomes, where return to the state immediately prior is not possible). Barshes looked at 1,000 repeated simulations of 100,000 hypothetical diabetes patients with no current or historical DFU, over a period of five years in 1-month intervals. Each month, each “patient” would exist in one of six clinical states: no DFU, uninfected DFU, infected DFU, limb loss, healed DFU, and all-cause death. Based on available clinical data, the patients were stratified into low, moderate, and high risk, and transition probabilities for moving from state to state each month were assigned (for example, the chance of transitioning from no DFU to initial DFU event in moderate risk patients was 0.3%, while the chance of limb loss in undertreated DFU in high risk was 3.1%). Each of the simulations was run with transitions occurring over five years (60 months/transitions), and the outcome probabilities were pooled. Each outcome was assigned a monthly cost estimate (for example, the median monthly cost of a healed DFU was $45, infected DFU $12,955, and major limb amputation such as BKA $38,934). Remember, each of these costs were per case, not the total population.

By applying costs of both primary and secondary preventive measures to all levels of risk-presenting patients (low to high), cost thresholds, at which at least 90% of simulations demonstrated savings, were established. An example was a measure that decreased the occurrence of DFU by 10% (0.90 RR), costing $50 per person and would have greater than a 90% probability of reducing amputations (at almost $39K) in diabetes patients at a cost that is equal or even lower than the standard of care, compared to no preventive care. The same 10% reduction in moderate- to high-risk patients from preventive care costs $125 per patient, with increases in cost as risk reduction also increases, yet said costs are considerably less than the outcome of amputation. For the purpose of this discussion, these results have been simplified.

The lack of programs designed to prevent/eliminate DFUs is troubling, this in spite of the known impact these DFUs have on amputation requirements, increasing healthcare costs, and overall quality of life. The paucity of such programs, even in larger academic healthcare centers, may be related to the perception of a clear lack of economic benefit.  Studies have been few and far between, and prior Markov models have not demonstrated a potential for overall savings, where cost effectiveness has been shown. The difference in this study from past offerings is this one looked at differing degrees of effectiveness (risk reductions ranging from 5% to 25%), assigning costs to each and determining a likely cost threshold for determining the need for preventive measures. One important limitation stated by the authors was separating low-risk from moderate- to high-risk patients, which may cause those higher risk populations to lose favor due to increased costs of prevention. An examination of the overall population as a whole would have been warranted to help support better utilization of prevention of diabetic foot ulcers and subsequent complications. If little else, there is certainly a need to encourage preventive programs as a means to reduce these high costs of care.

Practice Pearls:

  • Resistance to diabetic foot ulcer prevention programs seems driven by the lack of perceived “return on investment” of such measures.
  • Markov models can be used to demonstrate reduction in risk of developing costly complications from DFUs
  • A clear cost benefit can also be demonstrated, where utilizing relatively low cost prevention can result in avoidance of significantly costlier events.


Barshes NR, Saedi S, Wrobel J, Kougias P, Kundakcioglu OE, Armstrong DG. A model to estimate cost-savings in diabetic foot ulcer prevention efforts. J Diabetes Complications. 2017;31(4):700-7. Epub 2017/01/21. doi: 10.1016/j.jdiacomp.2016.12.017. PubMed PMID: 28153676.


Mark T. Lawrence, RPh, PharmD Candidate, University of Colorado-Denver, School of Pharmacy NTPD