Risk Assessment and Lifetime Integrity Analysis of Common Metallic Implants

The projected life expectancies of orthopedic implants in modern times are often being

usurped. As patients require joint replacements/implants earlier and are living longer healthier lives,

the integrity of aged implant devices comes in to question. New hurdles in risk assessment are

presented due to this required longer integrity of the implant materials. The major player as far as a

materials integrity issue is resistance to corrosion fatigue as common biomedical alloys, Ti-6Al-4V

and SS316LM, are not entirely inert in the body’s chemistry. Risk assessment analysis based on

these parameters of material performance are coupled with analyses of risk attributed to materials

defects, manufacturing quality control oversight, failed implant replacement, etc. The implicative

analysis tool utilized is a probabilistic modeling software package called @Risk, which provides a

bridge over the gap between technical data and decision-making.

Most modeling exercises are purely deterministic, i.e., single value inputs are used to obtain

a single value outputs, and the inputs are manually adjusted to attain the output. However, there is

uncertainty in these inputs that can be accounted for by probabilistic modeling, such as what is

available by software packages like @Risk and Crystal Ball. By using probabilistic models rather

than deterministic models, the inputs are distributions rather than singe values, and the outputs are

also distributions. Furthermore, plotted data is now given the added value of becoming a distributed

range of values rather than a single line or trend, and this distributed range can be assigned

gradations to indicate probability or certainty. In addition, performing sensitivity analysis is greatly

simplified, so that the completed model can rank critical variables and quantify their impact on the

result, often with minimal adjustments to the model. Probabilistic modeling is essential to evaluating

risk in engineering problems because of the uncertain nature of many engineering inputs, and since

the uncertainty is captured in the output, the result is much more informative than just a single

number. Probabilistic modeling puts data in context and allows decision makers to instantly see

critical technical issues and use the information provided to make decisions about materials choice,

operations, finances, or management.