Health – Case Statements 2017-11-08T10:11:20+00:00

Case 1

A medical funder asked us to investigate the prescribing patterns of general practitioners with the aim of detecting anomalous prescribing patterns.

We built a model comparing the pairing of in practice procedures with prescribed medication using a technique similar to market basket analysis in a retail environment to analyse common and less common pairing of items. Each practitioner was compared with the average pattern of a peer group of practitioners with a similar patient profile as informed by a clustering algorithm.

By showing the anomalous paring to a medical practitioner, we were able to identify several instances of abuse where certain treatments indicated for rare conditions e.g. anaemia were used by certain doctors as a booster shot. The doctors in question were challenged and additional guidance issued by the funder.

Case 3

We were asked by a managed care provider insuring the lower cost options of a number of medical funders at a fixed fee (capitation rate) to investigate the reasons behind a deteriorating loss ratio. The insurer had moved from a profit to a loss making situation within a period of less than a year. After eliminating the usual causes for a deteriorating claims experience e.g. aging membership, changing chronic profile our model pointed to a higher than expected hospital prevalence rate with no apparent cause. We created a series of visual discovery dashboards using QlikView and scheduled a workshop with the client to review. Using this approach we were able to isolate the members who had hospital admissions and systematically displayed these members in terms of all the features they had in common i.e. location, scheme, admitting physician, reason for admission, employer, age, chronic profile.

When plotting the number of hospital admissions over broker, the answer was there for everyone to see – a small number of brokers had signed up individuals which had hospital admissions shortly after inception date. Armed with this knowledge and a follow-up investigation, the client was able to establish the brokers in question had approached old age home residents. The situation was quickly remedied by the client insisting of stricter underwriting and pre-existing condition checks by the underwriting department. A portion of the loss was recovered by the client from the scheme in terms of a failure in agreed
underwriting procedures.

Case 2

A medical funder asked us to help them establish a network of medical specialists. This task involved geo-locating and mapping all members to the closest provider by doctor type according to accessibility criteria specified by the scheme. One the candidate specialists where identified, we used past claims experience and comparative peer group analysis to analyse and establish each specialists billing pattern.

Using this information, the scheme could approach the specialists and secure preferential rates for its members resulting in significant downward pressure on schemes overall cost.

Case 4

A medical scheme administrator asked us to assist them in attaching a value to their claims assessment activities. By analysing the difference between the original claims amount submitted, the scheme’s standard rates and the reasons for claims not being settled at what was charged, we were able to accurately assess the savings brought about by the administrator’s intervention. The analysis was also extended to other service providers and the final report included the claims assessment costs of each service provider. This report was automated and now forms an important part of the administrators submission to the schemes they administer.

The intervention report was also prepared on historic data for a scheme that was taken over by the administrator allowing the administrator to objectively show the scheme the significant savings it had managed to achieve compared to the efforts of the previous administrator. It also allowed the administrator to review its assessment rules to find a better balance between saving the scheme money and avoiding charges to members in terms of co-payment. The report has also been shared with the regulatory body to support the renewal of the administrators managed care and other
licensing provisions.

Case 1

A medical funder asked us to investigate the prescribing patterns of general practitioners with the aim of detecting anomalous prescribing patterns.

We built a model comparing the pairing of in practice procedures with prescribed medication using a technique similar to market basket analysis in a retail environment to analyse common and less common pairing of items. Each practitioner was compared with the average pattern of a peer group of practitioners with a similar patient profile as informed by a clustering algorithm.

By showing the anomalous paring to a medical practitioner, we were able to identify several instances of abuse where certain treatments indicated for rare conditions e.g. anaemia were used by certain doctors as a booster shot. The doctors in question were challenged and additional guidance issued by the funder.

Case 2

A medical funder asked us to help them establish a network of medical specialists. This task involved geo-locating and mapping all members to the closest provider by doctor type according to accessibility criteria specified by the scheme. One the candidate specialists where identified, we used past claims experience and comparative peer group analysis to analyse and establish each specialists billing pattern.

Using this information, the scheme could approach the specialists and secure preferential rates for its members resulting in significant downward pressure on schemes overall cost.

Case 3

We were asked by a managed care provider insuring the lower cost options of a number of medical funders at a fixed fee (capitation rate) to investigate the reasons behind a deteriorating loss ratio. The insurer had moved from a profit to a loss making situation within a period of less than a year. After eliminating the usual causes for a deteriorating claims experience e.g. aging membership, changing chronic profile our model pointed to a higher than expected hospital prevalence rate with no apparent cause. We created a series of visual discovery dashboards using QlikView and scheduled a workshop with the client to review. Using this approach we were able to isolate the members who had hospital admissions and systematically displayed these members in terms of all the features they had in common i.e. location, scheme, admitting physician, reason for admission, employer, age, chronic profile.

When plotting the number of hospital admissions over broker, the answer was there for everyone to see – a small number of brokers had signed up individuals which had hospital admissions shortly after inception date. Armed with this knowledge and a follow-up investigation, the client was able to establish the brokers in question had approached old age home residents. The situation was quickly remedied by the client insisting of stricter underwriting and pre-existing condition checks by the underwriting department. A portion of the loss was recovered by the client from the scheme in terms of a failure in agreed
underwriting procedures.

Case 4

A medical scheme administrator asked us to assist them in attaching a value to their claims assessment activities. By analysing the difference between the original claims amount submitted, the scheme’s standard rates and the reasons for claims not being settled at what was charged, we were able to accurately assess the savings brought about by the administrator’s intervention. The analysis was also extended to other service providers and the final report included the claims assessment costs of each service provider. This report was automated and now forms an important part of the administrators submission to the schemes they administer.

The intervention report was also prepared on historic data for a scheme that was taken over by the administrator allowing the administrator to objectively show the scheme the significant savings it had managed to achieve compared to the efforts of the previous administrator. It also allowed the administrator to review its assessment rules to find a better balance between saving the scheme money and avoiding charges to members in terms of co-payment. The report has also been shared with the regulatory body to support the renewal of the administrators managed care and other
licensing provisions.