OR WAIT null SECS
© 2023 MJH Life Sciences™ and Pharmaceutical Commerce. All rights reserved.
Patients need support programs to overcome a variety of barriers to therapy. Uncertainty about copay is a barrier to therapy at the patient level. The requirement of step therapy creates a barrier for patients at the payer level. Benefits investigations provide an additional barrier to therapy at the payer level. Patient support programs play an important role in impacting patient health outcomes by helping patients gain access to therapy, persist on therapy, and adhere to treatment guidelines as prescribed by their healthcare physician (HCP). Optimization of patient support programs is key to ensuring that the greatest assistance is provided for the largest number of patients.
Assessing the efficacy of patient support programs is the first step to optimizing their resource allocation and program design for maximized patient benefit. The efficacy of a support program is assessed based on the program's goal. The overall goal of support programs is to assist patients in receiving the therapy they need. As a result, the successes of support programs are assessed by their effectiveness in driving patient-centric metrics.
Patient support programs intend to enhance patient health outcomes. These outcomes are often not directly measurable because of data limitations and other considerations (privacy, for instance). As a result, novel patient outcome-focused key performance indicators (KPIs) are needed to measure the true efficacy of support programs. These KPIs, such as treatment fulfillment, patient persistence, and therapy adherence, serve as important gauges of a program’s success in enhancing patient health outcomes.
Analysis of patient support programs comes with its own set of problems from the onset. Limitations in data restrict how much can be accounted for during the analysis. Because patient information is required, privacy restrictions become a major data limiting factor. The need for multiple data sources breeds data disagreements, which causes uncertainty about the data. These data issues pose difficulties in initiating the study of support programs and their effect on patient therapy.
Upon circumventing and solving the data issues, there still are difficulties in the analysis itself. Changes in the market influence the observed result and affect them inconsistently over time. Changes in payer behavior add another level of complexity to the analysis. Limitations in the study period can create potentially misleading results for patients at the edge of the study window. These factors further confound the study needed to measure and optimize patient support programs.
Individually, each KPI only tells part of the patient's support story, and alone they can be misleading. For example, if a support program increases patient persistence, but reduces therapy adherence, then the net days on therapy may be unchanged. By considering persistence and adherence together, it becomes possible to establish how many additional days of therapy a patient support program truly bestows on patients. Together, patient persistence and therapy adherence can assess the additional days of therapy gained by program-affected patients, but they still only account for the impact on existing patients.
Assessing the effect of a support program on treatment fulfillment establishes the number of new patients brought on to therapy. The number of days on therapy of the new patients is established with the expected days on therapy, which is found by analyzing the support program's effect on persistence and adherence. Taken together, the total number of therapy days gained by an individual support program for the affected patient population is established. Knowing the true overall effect of a patient support program on the total days of therapy it gains is the first step to rating and ranking support programs by effectiveness. Support programs that are less effective in an overall sense are identified and can be curtailed to allow for better resource allocation and increased effort focused on the highest achieving programs.
While simply electing to expand the patient support programs that yield the greatest increase in total patient therapy days is a good holistic strategy for optimizing support programs, it may not be the best way to proceed depending on the current product's needs. The isolation of individual effects on each KPI for each support program enables a targeted approach to patient support depending on the goal. Different products may have different needs and thus require a different approach. Optimization of support programs is done by finding what works best for a given product.
Newer products may have a greater initial access barrier compared to better established ones. For such cases, focusing on expanding patient support programs that are most effective in increasing treatment fulfillment would be a better use of resources. As a result, there is less focus on patient persistence or therapy adherence-driven support programs. Conversely, an older product with fewer barriers to initial access would benefit from support programs that are more effective for increasing patient persistence and therapy adherence. A targeted approach to optimization of patient support programs is best for addressing different situations as they change over time; all of this only becomes possible with an accurate assessment of a patient support program's effect on patient-driven KPIs.
Overall, program efficacy is important, but it is not the be-all-end-all. Nuances in interactions between programs are of great significance when establishing the true value of a support program. Even programs with lesser efficacy may have value under the right conditions. Two or more support programs being used in tandem may have a greater effect on various KPIs than expected based on their individual effectiveness; this synergistic effect between programs can be used to find unforeseen value in otherwise less effective programs.
Identifying synergistic effects between programs allows for strategic planning of program combinations to ensure the greatest effect on the patient outcome without sacrificing resources. Support programs that are most effective when paired can be combined into a single super-program or distributed in a manner that promotes their use in tandem. Doing so ensures that both programs are used to their maximum effect. Ultimately, optimization in support program resource allocation and program combination is for the benefit of the patient, as it ensures that the most support possible is provided.
The first issue that arises with any analysis of patient outcomes is gathering and consolidating data. Specific information is needed about not only the patient's therapy but also the patient themselves. Privacy restrictions limit the level of detail of the patient information available and restrict the scope of its uses. While there are limitations on how much information can be collected, it becomes possible to build a more complete patient profile when information is consolidated from multiple sources. Information that is missing from one source may be present in another, but this does lead to another issue.
An issue with pooling data sources and data collection overall is reconciling different data. Patient and market information comes from a variety of different sources. Disagreements between data sources and differences in assumptions create immensely difficult situations for reconciling different pieces of data. These issues create ambiguity in information and raise concerns about data reliability.
There are two major strategies for dealing with disagreements between data sources, both of which need to be used depending on the situation in which the disagreement is presented. The first method is to build a consensus by a majority. If more data sources support a given value, then it is more likely to be correct. The other method is to simply use the most recent value, as it may be based on the most up-to-date information. Data collection and reconciliation are difficult tasks but are requisite for the analysis of patient support program effectiveness.
Changes in data over the course of the study period supply another level of complexity to the analysis. The changes in the market over time can favor or disfavor certain support programs depending on the point in time considered. The sway of market events can confound the observed effects of a support program. For example, support programs launched at the start of the COVID-19 pandemic may appear to be less effective because of a negative effect on the market caused by the pandemic itself. Studying variance in the patient outcome of program-unaffected patients over the course of the study period allows for identifying market events that may affect observed support program impacts at different points in time. Additionally, the market study helps further identify market trends and events that were previously unconsidered.
Other confounding factors, such as changes in insurance coverage, further complicate the analysis of support program effectiveness. The inclusion of information such as insurance provider and market access in the analysis allows for a controlled study of support programs. Changes in payer behavior and the market, in general, are critical factors that must be considered and accounted for during the study of patient support program impact.
The effects of patient support programs are not necessarily immediately felt. When analyzing patient support programs over a restricted period, delays in program impact can drastically change the observed effect on patients closer to the beginning or end of the study period. Two methods are used to account for two different confounding events. The first method is the use of a look-back period. By considering program exposure that happened to patients prior to the study period, the risk of missing the impact of a support program that was not observed in the study period is greatly reduced. Additionally, the use of a look-forward period allows for more accurate observation of patient outcomes.
The difference between using look-forward/back periods and simply expanding the study period is in the patients that are considered for the study. Only patients that are initiated during the study period are considered for the study. Additionally, while their ultimate outcome is found in the look-forward period, none of the support programs during that time are considered, as they may not have enough time to take effect. The use of a look-back period is important for proper support program attribution, and the use of a look-forward period is imperative for correct patient outcome assessment.
Patient support programs cannot be developed in isolation and become useful only when they are developed with a comprehensive assessment of program intent, design, operations, and data considerations. Moreover, these programs often interact with each other either by design or by accident, further strengthening the need for comprehensive understanding. We move towards efficiency by looking at the individual and synergistic effects between programs, strengthening and expanding them. Conversely, programs with redundant effects can be curtailed to allow for optimal utilization of resources and to avoid waste.
Another opportunity that can be pursued is targeting individual KPIs for support depending on the product and its current place in the market. Newer products may benefit more from a targeted approach that focuses more on bringing new patients to therapy rather than expanding the therapy of existing patients. Conversely, older products may need to focus on enhancing support for existing patients.
Analysis of support programs can be problematic due to data constraints and other confounding factors that add complexity. Differences in information between data sources cause ambiguity in results. Disagreements between data sources can be resolved by forming a majority consensus or using the most up-to-date information.
Changes in payer behavior or the market in general add extra layers of complexity to correct attribution of support program effects to patient outcomes. The study of variations in the market overall helps identify market factors that need to be considered. Inclusion of payer information and payer market access information in patient outcome analysis is needed to account for patient-level events and status. The inclusion of support program exposure information that occurred prior to the period of study is necessary to identify program-exposed patients correctly. Similarly, it is also necessary to include information on patient outcomes that occur after the study period to assess the program's effect on patient outcomes correctly.
Ramesh Krishnan, Ph.D., is a Principal and Jacob Maman is an Associate, both in the Decision Science team at Axtria.