Getting the Most Out of Longitudinal Patient Data

Anonymous patient-level data (APLD) is data collected in real time from an individual patient. There has been an increasing interest in patient-level data, as researchers, healthcare providers, and pharmaceutical companies are realizing the potential of creating better comparisons of effective treatment outcomes by analyzing long-term data that represent individual patient-based experiences.

What is anonymous longitudinal patient data (APLD)?

Anonymous longitudinal patient data (APLD) is patient-level data that can be used by pharmaceutical companies to track patient treatments over time, or to answer specific business questions.

Where does longitudinal patient data come from?

APLD can be obtained from various sources, including electronic medical records, pharmacy records, physician surveys, hospitalized patient billings, medical claims data, and patient diaries. The common characteristic of this data is that it reflects actual patient behavior in real time and not after-the-fact. Therefore, there is no risk of interviewer bias or recall errors.

APLD is aggregated and sold by companies like Symphony (PRA Health), IQVIA (formerly QuintilesIMS), Truven (IBM), Flatiron, and Optum. Also, health data exchanges, like Health Verity are making a push into the APLD market.

Where does APLD Fall Short?

Inconsistent data collection process

With their extremely rigorous schedules, physicians do not always take the time to record data in a comprehensive manner. As a result, inconsistencies and gaps in data are introduced into APLD. These inconsistencies and/or gaps in data are a major challenge to robust analytics.

Varying availability of historical data

APLD is a collection of real-time ongoing data. As historical data is a different concept, there is varying availability of historical data. Data consistency and formatting can be a challenge when used for analytical purposes, as fields can often be missing or improperly labeled, and formats vary by system and institution. 

How can healthcare organizations get the most out of APLD?

By using efficient longitudinal patient level data analytics, healthcare organizations can benefit really well from APLD. The following points highlight the different ways healthcare organizations can get the most out of APLD.

APLD allows for extensive anonymous tracking over time

APLD allows pharmaceutical organizations to track patients anonymously over a specified desired length of time. In Pharma and healthcare, more data usually amounts to fewer mistakes. Therefore, pharmaceutical and healthcare organizations view extensive anonymous tracking as an advantage because they can learn more about how patients respond to treatments over time. Moreover, companies can also use APLD to evaluate which treatments are most effective in improving a patient’s medical condition.

APLD can answer business questions

In order to stay ahead of the competition, pharmaceutical and healthcare organizations need to understand the market. APLD offers companies valuable data that can help to answer important business questions about the market, including the following:

What types of patients are being treated with products in the organization’s target market? Understanding patient characteristics and demographics such as age, gender, ethnicity, and geography can help companies strategize and target the right people.

What are the sources for getting new patients? APLD can identify the most popular sources for acquiring patients, for example, analysis of APLD can show whether patients are brand new, switching from other treatments, or repeating customers.

Are patients following the recommendations of their physicians? Drug adherence and patient compliance are very critical for the overall effectiveness of a drug. Healthcare organizations can use APLD to evaluate drug adherence. Additionally, APLD analytics can identify whether or not a specific demographic appears less compliant than others.

Why (and how frequently) is a patient's therapy changing? Strong analytics tools can help healthcare organizations identify why, and perhaps how frequently, a patient’s treatment is being changed by a healthcare provider.

Which products are losing/gaining prescriptions (and from where)? APLD offers the potential for real-time comparisons between prescription options. Therefore, analysts can figure out what prescriptions are losing/gaining patients.

What are the reasons for switching a patient to a different treatment? APLD offers the potential for going beyond the data to analyze the underlying causes for switching patients to alternative treatments. For example, companies can see when a prescription was denied due to cost, or where incentives impact prescriptions (i.e. the effectiveness of discount cards/coupons).

What are the market dynamics? When working with multi-country data, companies should be aware that different countries have unique ways to collect data and also have different healthcare systems. APLD can provide readily-available information on differences in market dynamics so that companies can specify different targets based on location.

Advanced data tools can use APLD to identify and fix gaps and formatting inconsistencies in data

By using advanced data science tools, healthcare organizations can identify and fix gaps and formatting inconsistencies in APLD, and hence, make the data more fit for analysis. Further analysis can then be used to educate patients and improve the quality of medical care. Further analysis is beneficial for improving the quality of research and educating physicians and other healthcare providers.

Leverage advanced data transformation techniques to reformat data on the fly during analysis

In recent years, there has been a heightened awareness that performing data analysis on individual patient data is beneficial.1 Tracking APLD for an extended period of time allows healthcare organizations to explore using the data intelligently, for instance, healthcare organizations can utilize evidence-based content to figure out immediate patient-specific needs as APLD is readily available compared with “aggregate” data. A good example of advanced data transformation techniques from APLD can be seen in a 2015 study that performed a meta-analysis of APLD to directly compare the differences in standard-of-care prostate cancer treatments.2

Due to the use of different data sources for individual patients, healthcare organizations need to also leverage tools that have the power to standardize the data across the board. One useful emerging model for converting datasets into a standardized form is the OMOP Common Data Model (CDM). With the OMOP CDM, healthcare organizations or pharma firms can take disparate observational databases and transform them into a common format (data model), and then move forward in performing systematic analyses.

Overall, APLD is very beneficial because it is not subject to influence as clinical trials are. Therefore, APLD analytics can be leveraged to show real world performance of drugs. APLD analytics can also make comparisons to competitor drugs in real-time to influence prescription choices and formulary coverage.

Use of APLD can support insight for all phases of product lifecycle

Since APLD is readily available and can be used for pharmaceutical data analysis, it can support all phases of the product lifecycle – discovery, development, and commercialization.

longitudinal patient data


Benefits of partnering with a healthcare-specific data science software expert that specializes in longitudinal patient data and offers analytics consultation

Partnering with a healthcare-specific data science software expert in Pharma analytics companies can help healthcare organizations understand the various uses of APLD and how to leverage it efficiently to create better overall patient experiences.

At RxDataScience, our apps contain advanced analytics and we specialize in mapping patient journeys (or patient level longitudinal data) over a long period of time. In addition to our micro-apps, we offer an innovative analytics-as-a-service platform (on the cloud or on premises). We also see an opportunity to achieve OMOP standardization via an API/data trans approach (versus reorganizing the actual databases); our API/data trans approach presents a lower lift and is more economical.

References

  1. Riley RD, Lambert PC, Abo-Zaid G. Meta-analysis of individual participant data: rationale, conduct, and reporting. BMJ. 2010;340. doi:10.1136/bmj.c221.
  2. Geifman N, Butte A. A patient-level data meta-analysis of standard-of-care treatments from eight prostate cancer clinical trials. Sci Data. 2016;3:160027.