11.

CASE STUDY: HEALTHCARE ANALYTICAL DASHBOARDS USING TABLEAU

11.1 Problem Definition:

The Quality of the healthcare data is enforced as part of Patient Protection and Affordable Care Act and to

measure the quality of the data that is used to analyze outcomes. Currently the health plans dont have an

easy way to analyze and understand quality of the health care data received from the trading partners.

Healthcare data provided by Wily Fox Technologies is used for the reports. This data has been

processed and analyzed for quality through a proprietary healthcare analysis system. It does not contain

any protected health information (PHI). The data provided is in excel format.

The reason for measuring the quality of data is to try to improve it and to perform benchmark

analysis to get a clear understanding of the quality of the dataset. The quality metrics can be used to drive

data quality improvement efforts, to ensure that analysis of healthcare related data sets is

accurate. Improved data quality supports improved analysis, management, and policy setting. It will also

provide a consistent basis for withhold and incentive programs to facilitate the shift to the pay-for-

performance delivery model.

11.2 Objective:

As mentioned above measuring Health Care data quality is very important. Objective of this Case study is

to see how Tableau software can measure data quality and answer following questions by drill down

dashboards.

How many trading partners are up to standards (which we categorize by Pass/Fail)?

What is average percentage of encounters submitted in days?

 Details of each trading partner and its performance.

What is the overall performance of trading partners?

Measure of overall Data quality.

Following Dashboards are created to analyze health care quality of the trading partner submitted data.

Duplicate Encounters Dashboard

Dental Encounter Lag Time Dashboard

Institutional Encounter Lag Time Dashboard

Professional Encounter Lag Time Dashboard

Pharmacy Encounter Lag Time Dashboard

Turn Around Time Dashboard

Quality Measure Summary Dashboard (Main Dashboard)

11.3 Background:

The Agency for Healthcare Research and Quality (AHRQ) defines quality health care [7] as doing

the right thing at the right time, in the right way, for the right person and having the best possible

results [7]. In health care, the difference between good and poor quality can literally mean the

difference between life and death.

The Institute of Medicine defines healthcare quality by six components:

Effective Providing services based on scientific knowledge to all who could benefit.

Safety Avoiding injuries to patients from the care that is intended to help them.

Patient Centered Providing care that is respectful of and responsive to individual preferences,

needs and values and ensuring that patient values guide all clinical decisions.
 Timely - Reducing waits and sometimes-harmful delays for both those who receive and those

who give care.

Efficient - Avoiding waste, including waste of equipment, supplies, ideas, and energy.

Equitable - Providing care that does not vary in quality because of personal characteristics such as

gender, ethnicity, geographic location, and SES [7].

To lower health care costs and provide quality, government has issued Affordable Care Act. This

act provides better health protection by filling current gaps in coverage, hold insurance companies

accountable, lower health care costs, guarantees more choice, expand coverage and enhance the

quality of care for all.

11.4 Healthcare Terminology:

In simple terms,

Trading Partners are different Health care providers.

Encounters are records submitted by trading partner to the government.

As we have an idea on what are trading partners and encounters, we will see few other terminologies and

its measures.

Duplicate Encounters

Duplicate encounters are encounters submitted by trading partners that are duplicates of previously

submitted encounters.

The expected outcomes are measured by transaction type:

- Less than 7% of Institutional encounter records are duplicates of already accepted encounter

records.
 - Less than 7% of Professional encounter records are duplicates of already accepted encounter

records.

- Less than 4% of Pharmacy encounter records are duplicates of already accepted encounter

records.

If any of these thresholds are exceeded, this measure fails. If all are met, this measure passes.

This measure reviews the percentage of encounter records submitted by a trading partner that are

duplicates of previously submitted encounter records.

Lag Time

Lag time is the length of time between the service date and the submission date that the encounter was

received by the trading partner. The service data is the actual date that them member received healthcare

services. The submission date is the date that the enforcement agency received the file containing the

itemized services.

The four lag time categories are:

- 60% records where lag time is zero to 90 days

- 80% records where lag time is 91 to 180 days

- 95% records where lag time is 181 to 365 days

- 5% records where lag time is greater than 365 days

This measure addresses the lag time for submitting encounter data. Lag time is the time, in days, between

the date of service and the date of submission by a trading partner. If the percentage of records falls below

the percentage in the lag time categories, then the trading partner will not pass the quality measure.

In this project we deal with four different lag times for the following healthcare data categories:

- Institutional services for members where individuals require a prolonged period.

 - Professional the most common services that members receive (outpatient)

- Pharmacy Pharmacutical related services (prescription drugs)

- Dental Outpatient dental care services.

Turn-Around Time

Turnaround time is the time between when a denial encounter was submitted and when the denial

encounter data has been corrected.

- 50% of denials should be corrected within 15 days

- 80% of denials should be corrected within 30 days

- 95% of denials should be corrected within 60 days

Greater than a 5% descrepancy in any of the above categories is a failure to meet the measure.

This measure addresses how quickly denied encounter records are corrected and resubmitted by a trading

partner. The time between a denial and the correction and resubmission of corrected data is the

turnaround time. This measure analyzes the percentage of corrections by turnaround time categories.

Denied encounters have failed the data quality edit process and plans must correct them in a timely

fashion.

After understanding healthcare terminology, it is easy to get into creation of visualizations with the data

using tableau.

11.5 Solution:

Tableau software is used to create following dashboards each of which will able to analyze the quality of

data submitted by the trading partners. This Dashboards will give the user a better visual representation of

the data quality which can be used to make more appropriate decisions and also to contact each trading

partner to check the issues with their data quality.

Dashboard 1: Dental Encounters Lag Time Dashboard

In this dashboard, Lag time of Dental encounters is analyzed. Dental lag time is the length of time

between the service date and the submission date that the encounter was received by the trading partner

for Dental Encounters.This measure addresses the lag time for submitting encounter data as a

percentage.This dashboard contains,

- Number of Trading Partners that Pass or Fail based on the four categories of Dental lag time

in zero to 90 days, lag time in 91 to 180 days, lag time in 181 to 365 days, lag time in greater than

365 days of thresholds.

- Average percentage of encounters in four lag categories.

- Detail information of each trading partner and its associated Result, percent lag in zero to 90

days, lag time in 91 to 180 days, lag time in 181 to 365 days, lag time in greater than 365 days.

Figure 11-1: Dental Encounters Lag Time Dashboard

Figure 11-2: Dental Encounters Lag Time Trading Partners Passed

Figure 11-3: Dental Encounters Lag Time Trading Partners Failed

Dashboard 2: Duplicate Encounters Dashboard

In this dashboard, duplicate encounter quality of data for Institution, Pharmacy and Professional

encounters is analyzed .This report will help the users to identify how many duplicate encounters are

submitted as a percentage of total encounters. This report displays percentage of data duplicate vs non-

duplicate data received is displayed. This report is an interactive report where users can drill down into

the specific trading partner/ type of encounter to get more information. This dashboard contains,

- Number of Trading Partners that Pass or Fail based on the thresholds.

- Average duplicate encounters percent for Institutional, Pharmacy and Professional.

- Detail information of each Trading Partner and associated Institutional duplicate percent,

Pharmacy duplicate percent, Professional duplicate percent.

Figure 11-4: Duplicate Encounters Dashboard

Figure 11-5: Duplicate Encounters Trading Partners Passed

Figure 11-6: Duplicate Encounters Trading Partners Failed

Dashboard 3: Institutional Encounter Lag Time Dashboard

In this dashboard, Lag time of Institutional encounters is analyzed. Institutional lag time is the length of

time between the service date and the submission date that the encounter was received by the trading

partner for Institutional Encounters.This measure addresses the lag time for submitting encounter data as

a percentage. This dashboard contains,

- Number of Trading Partners that Pass or Fail based on the four categories of Institutional lag

time in zero to 90 days, lag time in 91 to 180 days, lag time in 181 to 365 days, lag time in greater

than 365 days of thresholds.

- Average percentage of encounters in four lag categories.

- Detail information of each trading partner and its associated Result, percent lag in zero to 90

days, lag time in 91 to 180 days, lag time in 181 to 365 days, lag time in greater than 365 days.

Figure 11-7: Institutional Encounters Lag Time Dashboard

Figure 11-8: Institutional Encounters Lag Time Trading Partners Passed

Figure 11-9: Institutional Encounters Lag Time Trading Partners Failed

Dashboard 4: Professional Encounter Lag Time Dashboard

In this dashboard, Lag time of Professional encounters is analyzed. Professional lag time is the length of

time between the service date and the submission date that the encounter was received by the trading

partner for Professional Encounters.This measure addresses the lag time for submitting encounter data as

a percentage. This dashboard contains,

- Number of Trading Partners that Pass or Fail based on the four categories of Professional lag

time in zero to 90 days, lag time in 91 to 180 days, lag time in 181 to 365 days, lag time in greater

than 365 days of thresholds.

- Average percentage of encounters in four lag categories.

- Detail information of each trading partner and its associated result, percent lag in zero to 90 days,

lag time in 91 to 180 days, lag time in 181 to 365 days, lag time in greater than 365 days.

Figure 11-10: Professional Encounters Lag Time Dashboard

Figure 11-11: Professional Encounters Lag Time Trading Partners Passed

Figure 11-12: Professional Encounters Lag Time Trading Partners Failed

Dashboard 5: Pharmacy Encounter Lag Time Dashboard

In this dashboard, Lag time of Pharmacy encounters is analyzed. Pharmacy lag time is the length of time

between the service date and the submission date that the encounter was received by the trading partner

for Pharmacy Encounters.This measure addresses the lag time for submitting encounter data as a

percentage. This dashboard contains,

- Number of Trading Partners that Pass or Fail based on the four categories of Pharmacy lag

time in zero to 90 days, lag time in 91 to 180 days, lag time in 181 to 365 days, lag time in greater

than 365 days of thresholds.

- Average percentage of encounters in four lag categories.

- Detail information of each trading partner and its associated result, percent lag in zero to 90 days,

lag time in 91 to 180 days, lag time in 181 to 365 days, lag time in greater than 365 days.

Figure 11-13: Pharmacy Encounters Lag Time Dashboard

Figure 11-14: Pharmacy Encounters Lag Time Trading Partners Passed

Figure 11-15: Pharmacy Encounters Lag Time Trading Partners Failed

Dashboard 6: Turn-Around Time Dashboard

Turnaround time is the time between when a denial encounter was submitted and when the denial

encounter data has been corrected This measure addresses how quickly denied encounter records are

corrected by a trading partner. The time between a denial and the correction and the corrected data is

succesfully submitted is the turnaround time. This measure analyzes the percentage of corrections by

turnaround time categories. Denied encounters have failed the data quality edit process and plans must

correct them for resubmission.This dashboard contains,

- Number of Trading Partners that Pass or Fail based on the three categories of turn-around

time, they are 15 days or less, time in 16 to 30 days, time in 31 to 60 days, of thresholds.

- Average percentage of encounters in three time slot categories.

- Detail information of each trading partner and its associated result, turn around time percent for

less than 15 days, 16 to 30 days, 31 to 60 days, and total percent denied.

Figure 11-16: Turn-Around Time Dashboard

Figure 11-17: Turn-Around Time Dashboard Trading Partners Passed

Figure 11-18: Turn-Around Time Dashboard Trading Partners Failed

Dashboard 7: Quality Measure Summary Dashboard

The Main Dashboard is a centralised dashboard which has links to all quality measure dashboards. This

Dashboard displays overall quality performance of the encounters data for all the trading partners. This

dashbaord has three sections.

Section 1 Trading Partner Performance:

This section gives a quick glance on performance summary of trading partners in numerical form.

Section 2 Data Quality:

Data Quality depicts the overall percent of approved encounters, total number of encounters submitted

and number of encounters approved and denied.

The color code, green number of approved encounters, red number of denied encounters.

Section 3 Quality measures:

This section provides individual quality measure performance in terms of percent for Pass and Fail. Each

quality measure is provided with link to the respective quality measure dashboard for greater detail. When

an enduser uses a mouse click on a quality measure name, they are redirected to that respective quality

measure dashboard.
 Figure 11-19: Quality Measures Summary

11.6 Conclusion:

The dashboards created here facilitates improved analysis, data management and policy setting. It

facilitates nested reporting. Its Interactive drill down features helps to understand the details of individual

trading partner performance and encounters data. The dashboard, reports and data extracts created in this

project can be used at various levels of an organization, from executives to staff that work with trading

partners on a daily basis.

Working on this project has helped me to understand the importance of data mining concepts, performing

analytics on the data available, along with researching and learning Tableau desktop tool and its

features. This also helped me improve my problem solving techniques. In addition, it facilitated me to

acquire good knowledge on healthcare domain. Apart from me, it also helps readers understand the usage

of tableau desktop tool.