October 16, 2014
A Better Way Than Predictive Modeling
by Karen Wolfe
Issues indicating a workers' comp claim is becoming a major problem can be addressed in real time -- if the right systems are in place.
Even though it’s obvious that early intervention drives better outcomes, few systems are in place to guarantee early intervention into problematic workers’ compensation claims.
One widely acclaimed effort to identify troublesome claims early is predictive modeling, which conducts elaborate analysis of data using advanced mathematical methods. But finding those claims remains a guessing game. Not every future problematic claim will be tagged, because many do not meet the modeling criteria. Predictive modeling can be helpful but is imperfect and costly.
A more practical methodology for identifying costly claims is to monitor the data on a concurrent basis to uncover dicey conditions in them. Rather than predicting which claims will be risky, the conditions that portend risk and cost in claims are isolated and then identified as they occur.
Technology is used to find the claims that bear those conditions whenever they occur throughout the course of the claim. All claims are monitored, so none are missed. No guesswork is involved.
The data monitoring approach is powerful, but, as with predictive modeling, only when the next step is taken. Organizations that undertake a process for identifying risky claims early stand to lose the entire benefit unless they also structure procedures for intervention. The appropriate persons must be notified about problematic claims immediately, and those persons must carry out the recommended procedures.
Whether the alert recipients are claims adjusters, medical case managers or someone else, they must follow intervention procedures. Using the data monitoring approach, each condition that is sought in the data should be associated with a prescribed intervention procedure. Follow-up procedures should be specific so that analyses can be made of their effects.
For instance, when the condition identified in a claim is the third prescription for a Schedule II drug, the system might be set to alert the medical director. A standard method of intervention or approach for intervening with the treating doctor is then followed.
If medical treatment is extended past a designated point for low back strain, an alert would be sent to the claims adjuster, whose procedure is to engage a medical case manager to investigate. The investigation procedure is standardized.
The important thing is that intervention procedures are analyzed in advance and clearly stated so the actions are consistent across the organization and over time, regardless of who actually carries out the intervention.
Medical case management as a strategy has been undervalued because each intervention was handled individually, meaning that results could not be measured. But it’s possible to standardize and categorize tactics so outcomes can be measured and compared.
Claims adjusters sometimes fail to refer cases to medical case management for a variety of reasons. When early intervention procedures are standardized, the referral is automatically made by the system, thereby eliminating the burden of referral for claims adjusters. Efficiency is good.
Simply stated, early intervention is more effective than late intervention. The problems have not yet morphed into catastrophes and are usually easier to solve. When systems and procedures are established to automate problem claim identification and follow up procedures, best results can be pinpointed. The approach that produced those results can be continued. Approaches that produce lesser results can be modified.
The upward spiral of quality improvement is continuous.