Project Examples

Prior to a planned company merger, this client required – with a view to ensuring adherence to competition law – an informed estimate of the future joint market shares by region.

I set up a database containing all market share data for both companies, and then I visualized regional market shares (using the GFK-Regiograph®) and identified regions with potentially critical market shares. Next, I created a software tool to simulate various market coverage scenarios. This enabled both companies to consider a range of different business constellations in their negotiations and work out an appropriate solution.

A hospital’s cardiology department recorded a dramatic drop in the number of patients diagnosed with acute coronary syndrome, as well as the related treatments (such as coronary angiography and percutaneous transluminal coronary angioplasty, or PTCA). This raised the question of whether this was due to changes in the population structure, in the collaboration with referring physicians, or in the competitive environment.

A target/actual comparison of case numbers helped shed some light on the issue. First, I developed a forecasting model to determine what the “normal” case numbers would be, considering the demographic and micro-geographic structures of the hospital’s catchment area. I fed the model with primary data from German hospitals’ quality records (published by INEK) and from regional statistics (published by the state-level statistical offices). By comparing the forecast and actual figures – the latter are always available at hospitals as a standardized data set – I found that the drop in patient numbers had been sharpest in one particular zip code area. So the solution to the puzzle was obvious: A new cardiac cath lab had recently opened in that same area.

Another hospital was facing the reverse issue: The demand for cardiac surgery had risen notably. Again using the approach described above, I first verified that case numbers were actually above average. By comparing them with regional social structure data (from community statistics), I was able to identify a plausible reason: The hospital was located in a structurally weak region with high unemployment rates, a setting statistically associated with an increased incidence of coronary heart conditions (and several other diseases). With these insights factored into the equation, case numbers turned out to be within the normal range.

Every four to five years, the Italian Society of Invasive Cardiology (GISE) conducts a survey among its members to learn about the current availability of cardiologic services in Italy, and thus to be able to initiate improvements. Analytic Services was hired to do the survey (online) and evaluate its results.

After finding a skilled provider to build the platform (www.rete-ima.it), I conducted the survey at two levels: One questionnaire addressed the managers of hospital cardiology departments as well as the GISE officers in charge of provinces and regions; another addressed ambulance service officers at the province level. The Society had provided me with a database, which I enriched with statistical data on the population structure as well as micro-geographic information such as geo-coded hospital sites and distances to patients’ residences. The result was a quite precise picture of interventional cardiolo­gy coverage in Italy, which revealed striking differences between the country’s north and south, and thus offered a valuable decision basis for regional health policy.

A supermarket chain planned to modernize some of its stores, giving highest priority to those whose high-end assortments and customer profiles best fit the company’s new strategy.

Besides the usual market research data, I also worked with point-of-sales figures, as they provided valuable information on the sales per product group and thus the demand structure. After condensing the data to create a consistent basis for analysis, I worked out an approach based on multivariate analyses (such as factor and cluster analyses). It revealed striking correlations in the demand for certain product groups; at the same time, demand levels for those products differed greatly between branch stores, depending on their location and profile. Given these insights, I was able to define branch clusters serving different, relatively homogeneous customer groups and generating very different profit contributions. Two of these clusters had particular promise – they were the ideal candidates for store expansion and remodeling.

A pharmaceuticals company was about to introduce a new drug, and wished to learn more about its cost-value position in the German market.

I developed a simulation program to determine the cost and probable benefit of the drug, with the drop in transfusion rates used as an indicator of the latter. Primary data were available from clinical trials conducted in the U.S. and some European countries. To obtain meaningful results for the German market, I projected these data to the patient populations typically found here, as well as to the relevant subgroups, specifically the clients of statutory and private health insurances.

Based on this cost-benefit analysis, I was able to demonstrate the economic advantages of the new product compared to existing treatment regimes: a powerful argument for the client to use in negotiations with health insurances and hospitals.

In south-western Germany, several hospitals had formed a heart attack network. The goal was to ensure that myocardial infarction patients would receive fast and effective treatment in a joint intervention center. In the context of an initial situation assessment, the network wanted an analysis of the center’s cost-benefit ratio – specifically, how the quality of care (measured in quality-adjusted life years, or QUALY) and the costs of treatment had developed over the past three years.

Working closely with Dr. Steffen Wahler, MD, of St. Bernward GmbH in Hamburg (Germany), I started by analyzing the center’s records. For a comprehensive overview of care quality, I combined these data with regional hospital data on cardiac patients having undergone conservative treatment. To determine the approximate cost per surgical intervention, I searched the G-DRG-Browser of InEK (the German Institute for the Hospital Remuneration System).

By modeling all these data for the three years under study, I determined that treatment costs in the network had risen – but mainly because the center now handled more complex cases, rather than sending them to hospitals outside the network. In sum, the analysis confirmed that the network’s cost-benefit ratio was above market average.