7. Analytic & Vendor Mgmt - Health

Question 1

Define “data analytics” in the healthcare context

Answer 1

The process of inspecting, cleaning, transforming, interpreting, and modeling data to discover trends, patterns, and other information that can support benefit plan decisions & changes.

Goals:
1. Reduce costs
2. Improve clinical outcomes &/or the participant experience

Question 2

Define “predictive modeling” in the healthcare context

Answer 2

A statistical technique commonly used to forecast future behavior. It involves analyzing historical and current data to generate a model to forecast future outcomes. Can be used to quantify risk & costs for individuals & groups of folks enrolled in health plan.

Question 3

List 6 ways in which predictive modeling can be leveraged by health plans

Answer 3

1. Review a plan’s disease burden (health status) & how it will change over time
2. Stratify plan’s population by risk level to identify at-risk & catastrophic claimants for targeting disease mgmt & case mgmt, respectively
3. Identify risk factors likely to generate future plan costs that should be targeted w more intensive outreach, including finding at-risk individuals who, although may be low cost today, may generate significant costs in the future
4. Compare relative resource consumption* by groups for budgeting & underwriting forecasts
   *refers to how intensively plans use physician visits, hospital stays, & other member resources
5. Compare providers fairly, adjusting for differences in health risk among patient pop’ns. Such comparisons can be used to profile providers for utiliz review & quality of care
6. Analyze a medical mgmt program to see what the true savings are, as opposed to those that are regression to the mean *outcomes that are at least partly due to chance; refers to the phenomenon of “averaging out” in statistics

Question 4

Explain (8 pts) how health plan sponsors use data analytics & predictive modeling

Answer 4

1. Identify claims trends
2. Target high-risk users
3. Identify gaps in care
4. Steer patients to best providers
5. Measure vendor perf
6. Uncover cost-sharing strategies
7. Engage participants in their own care
8. Investigate waste, abuse, fraud

Question 5

List the 5 recommended steps plan sponsors should take to implement data analytics & predictive modeling tools

Answer 5

1. Determine who will perform the data analytics
2. Use data analytics & predictive modeling to identify & map the most prevalent clinical risk characteristics & associated costs in the plan pop’n
3. Establish a 3y health-mgmt strategy
4. Develop a formal participant comms strategy
5. Identify how plan participants will react to change

Question 6

There are 5 recommended steps plan sponsors should take to implement data analytics & predictive modeling tools. Describe the step “1. Determine who will perform the data analytics.”

Answer 6

Only the very largest plans have the capabilities to handle data analytics on their own. Most need to decide whether the analytics offered by their existing vendors are sufficient, or if they should outsource.

Question 7

There are 5 recommended steps plan sponsors should take to implement data analytics & predictive modeling tools. Describe the step “2. Use analytics/modeling to identify & map prevalent clinical risk & costs”

Answer 7

Plan sponsors should evaluate the programs in place to address clinical risk characteristics and associated costs in the plan population.

Question 8

There are 5 recommended steps plan sponsors should take to implement data analytics & predictive modeling tools. Describe the step “3. Establish a 3y health-mgmt strategy”

Answer 8

Strategy should have a budget, goals, and performance targets that increase over time.

E.g., improve wellness participation from 10% year 1, to 50% year 2, 75% year 3

Question 9

There are 5 recommended steps plan sponsors should take to implement data analytics & predictive modeling tools. Describe the step “4. Develop a formal participant comms strategy”

Answer 9

While data analytics can reveal the cost outliers to plan sponsors, effective comms can have an immediate, direct, positive impact

Question 10

There are 5 recommended steps plan sponsors should take to implement data analytics & predictive modeling tools. Describe the step “5. Identify how plan participants will react to change”

Answer 10

It’s important to remember that any changes a plan sponsor implements affects people directly.

Question 11

Identify 3 macro trends that presently impact the day-to-day operations of most HR teams, with details

Answer 11

1. Rising healthcare costs: Healthcare consistently outpaces inflation and makes up one of the largest line items in almost every company budget
2. Budgets are under stress: HR depts are being asked to do more with less. COVID accelerated the trend in many industries, and even in unaffected industries, uncertainty surrounding future variants has led companies to be more risk-averse & restrict spending
3. Workforce shifts: Attracting & retaining best talent is key to company growth & sustainability. Economic changes post-COVID have given EEs more choice than ever, leading to the largest talent acq and retention upheaval in 20+yrs

Question 12

Identify 4 types of analytics & the core questions they answer

Answer 12

1. Descriptive: what happened
2. Diagnostic: why it happened
3. Predictive: what might happen
4. Prescriptive: recommended actions

Question 13

Cute examples where the use of descriptive & diagnostic analytics can provide insights to plan sponsors re: emerging hc trends impacting their plans

Answer 13

Many situations arose re: pandemic.

COVID brought about many changes in the types of care folks sought, incl. rise in mental health utilization. By understanding change in EE needs like these, companies modified existing plans & searched for vendors to help them offer bens that served these expanding needs.

Deferred care also arose re: COVID. Many skipped/delayed care during pandemic. When health problems arise from lack from treatment, ER often see more expensive remedial treatments. To address deferred care, companies can take preemptive action to engage partners to boost annual screenings/physicals.

Pandemic saw rise in telehealth adoption. ERs altered plan design & comms to encourage uptake and drive savings for ER & EE.

In all 3 cases, descriptive & diagnostic analytics helped companies get a pic of what happened w their plans to address emerging trends.

Question 14

What 8 health plan trends have recently emerged that were documentable by descriptive & diagnostic analytics?

Answer 14

1. Shift away from single health plan offerings
2. More moderate health plan premium increase than originally projected at pandemic onset
3. ER absorbing a larger % of health premiums for their EEs
4. EEs participating in vol benefits, regardless of health plan choice
5. Increased participation in HDHPs
6. High HDHP selection level by Millennials
7. Increased enrollment & contributions to HSAs
8. Average total HSA contribs exceeding 60% of IRS statutory limits

Question 15

How can individual EEs utilize predictive analytics to take advantage of their ERsponsored health plans?

Answer 15

Decision support tools leverage health claims data from the previous year, allowing EE to predict OOP costs for each plan choice available to them.

EE can customize their expected usage based on what they expect to happen, e.g. birth of a child, to project & minimize OOP costs.

Question 16

Identify situations where ER use of predictive analytics could yield favorable financial outcomes for the ER

Answer 16

Measure the financial impact of plan changes across workforce, when considering plan provision changes. E.g., increasing ER and inpatient copay, increasing single deductible, adding family deductible.

Can identify what % of population would be affected by a change, & how individuals would be impacted.

Can estimate the ROI of a vendor program.

Question 17

How can ERs continuously learn from & improve their benefits programs by combining various analytic tools?

Answer 17

By combining data, computerized analysis algorithms, and automated actions.

E.g., in managing a Rx drug plan, plan sponsor could conduct ongoing analysis of Rx claims data, proactively identifying maintenance Rx with cheaper alternative, and automate process of getting the Rx switched.

Greater transparency would arise from putting drug cost info in the prescriber’s hands and take effort out of finding beneficial alternatives. Rx behavior change could create savings that complement the ER’s PBM/pharma consulting strategy, leading to a reduction in Rx spending.

Higher satisfaction can also occur for plan members bc they receive needed meds more affordably, keeps them healthy, improves medication adherence, saves money without effort on their part.

Question 18

What 6 criteria should a hc plan sponsor discuss with / ask a potential data analytics vendor?

Answer 18

1. Completeness of vision: Plan sponsors should look for vendors that can clearly outline how they have evolved to meet & anticipate industry needs.
2. Culture & values of senior leadership: The overall culture of a company starts at the top. Plan sponsors should get to know the senior leaders of the vendors they evaluate, and insist on meeting several members of the exec team. Is there alignment between vendor & sponsor leadership?
3. Ability to execute: Do the vendors in consideration have solid, referenced accounts, similar in size & demographics to plan sponsor’s own?
4. Tech adaptability & supportability: Underlying engineering & architecture of software. Plan sponsors, peel back the covers of vendor’s products; evaluate software engineering for modern design patterns
5. Total cost of ownership: Plan sponsors must understand total cost of vendor solution
6. Company viability: Will the vendor be around in 9y, the average lifespan of a significant IT investment?

Question 19

List the 10 key considerations for success when a hc org is implementing an analytics initiative:

Answer 19

1. Data modeling & analytic logic
2. Master reference/master data mgmt
3. Metadata repository
4. Managing white space data
5. Visualization layer
6. Security
7. Extract, transform, load (ETL)
8. Performance & utiliz metrics
9. Hardware, software infrastructure
10. Cultural change mgmt

Question 20

Explain this key consideration for success when a hc org is implementing an analytics initiative: “Data modeling & analytic logic”

Answer 20

Different vendors’ analytics solutions feature different data models. Which data model they use can have a significant effect on cost, scalability, & esp the adaptability of plan sponsor’s analytics solution to support new use cases

Question 21

Explain this key consideration for success when a hc org is implementing an analytics initiative: “Master reference/master data mgmt”

Answer 21

The ability to incorporate data from new and disparate sources into the plan sponsor’s analytics solution requires significant expertise in master data mgmt.

Question 22

Explain this key consideration for success when a hc org is implementing an analytics initiative: “Metadata repository”

Answer 22

Plan sponsors should look for a vendor that provides a tightly integrated, affordable, simple repository with its overall analytics solution

Question 23

Explain this key consideration for success when a hc org is implementing an analytics initiative: “Managing white space data”

Answer 23

Does the plan sponsor’s analytics solution offer a data collection alternative to the proliferation of desktop spreadsheets and databases that contain analytically important data? white space data is the data collected & stored in desktop spreadsheets and databases that is not being collected and managed in primary source systems, especially electronic medical records (EMRs), or it is being collected in clinical notes and must be manually abstracted for reporting & analysis. This desktop data fills in the missing “white space” of analytic info that is important to the org.

Question 24

Explain this key consideration for success when a hc org is implementing an analytics initiative: “Visualization layer”

Answer 24

The best analytics solutions include a bundled visualization tool - one that is both affordable and extensible if licensed for the entire org. However, the visualization layer is very volatile The leading solution today won’t be the leader tomorrow. Therefore, plan sponsors should look for an analytics vendor that can quickly and easily decouple the underlying data model and data content in the data warehouse from the visualization layer and swap the viz tool with a better alt when necessary.

Question 25

Explain this key consideration for success when a hc org is implementing an analytics initiative: “Security”

Answer 25

Privacy & security of patient data is paramount

Question 26

Explain this key consideration for success when a hc org is implementing an analytics initiative: “ETL”

Answer 26

A robust extract, transform, and load process - how analytics tech extracts data from source systems, applies the required transformations, and writes data into the target database - is fundamental to the success of the chosen solution

Question 27

Explain this key consideration for success when a hc org is implementing an analytics initiative: “Performance & utilization metrics”

Answer 27

Plan sponsors will need to generate metrics about who is using the system, how are they using it, and how well the system operates. Can the vendor’s solution track basic data about the environment, such as user access patterns, query response times, data access patterns, volumes of data, and data objects?

Question 28

Explain this key consideration for success when a hc org is implementing an analytics initiative: “Hardware & software infrastructure”

Answer 28

Viable and sustainable hardware & software vendors continue to change rapidly, esp with cloud-based products & svcs.

Question 29

Explain this key consideration for success when a hc org is implementing an analytics initiative: “Cultural change mgmt”

Answer 29

Tech is only part of the equation in creating a successful analytics program. A vendor’s solution must also include processes and real-world experience for helping the plan sponsor manage sustainable change in its analytics-driven org.

Question 30

Leaders in the analytics industry have developed a Healthcare Analytics Adoption Model that has 8 levels of analytics adoption that an org passes thru as it gains sophistication in using its data to drive improvement:

Answer 30

1. Enterprise Data Warehouse: Collecting and integrating the core data content
2. Standardized Vocabulary & Patient Registries: Relating and organizing the core data content
3. Automated Internal Reporting: Efficient, consistent production of reports & widespread availability in the org
4. Automated External Reporting: Efficient, consistent production of reports and adaptability to changing reqs
5. Waste and Care Variability Reduction: Reducing variability in care processes & focusing on internal optimization & waste reduction
6. Population Health Mgmt & Suggestive Analytics: Tailoring patient care based upon pop’n metrics. Fee-for-quality includes bundled per case payment
7. Clinical Risk Intervention & Predictive Analytics: Org processes for intervention are supported with predictive risk models. Fee-for-quality includes fixed per-capita payment.
8. Personalized Medicine & Prescriptive Analytics: Tailoring patient care based on pop’n outcomes & genetic data. Fee-for-quality rewards health maintenance.

Question 31

Explain why many orgs over time have moved from fully insured health plans to self-funded, ASO health plans

Answer 31

In a self-funded arrangement, the plan sponsor assumes the liability and risk associated with uncertain healthcare costs, but self-funded plans offer opportunities to lower costs & increase flexibility in plan design.

Self-funding ca give org tax benefits, better cash flow, and reduced admin costs.

Question 32

What is the relationship between ERISA & administrative claims audits of health plans?

Answer 32

Most carriers and TPAs do a good job of administering self-funded plans, but maintaining regular oversight of hc expenses is prudent and, in fact, a plan sponsor’s fiduciary responsibility. According to ERISA, it is the duty of plan trustees and other fiduciaries to act in the best interests of plan participants, including reducing claims expenses and ensuring the quality of administrative processes.

Question 33

Explain why administrative claims audits are more important in self-funded health plans than in insured plans

Answer 33

In the case of fully insured plans, most carriers will offer only a limited number of “standard” plan designs. But with a self-funded plan, sponsor has greater flexibility in design and admin. Bc the number of plan designs for a self-funded ASO client is potentially unlimited, complexity increases for carrier/TPA and creates potential for more admin mistakes.

Question 34

What are administrative claims audits?

Answer 34

A retrospective look at claims the carrier or TPA has paid, in order to identify possible processing errors, over- and under-payments. Such audits are conducted by accountants, independent hc cost-containment firms, and EE benefits consultants. Some large plans may have their own internal auditing departments.

Question 35

What are the % of claims found in most administrative claims audits that have been over/incorrectly paid, and is possibly recoverable?

Answer 35

1-3% of the total amount spent on claims annually is potentially over/incorrectly paid.

Question 36

Describe the audit timeframe for most administrative claims audits

Answer 36

Most agreements allow for only a 12- to 24-mo look back from the date the audit begins. While the ASO agrmt might allow for up to 24mos, the provider network contracts in place will usually allow for only 12mo when it comes to recovering claims that were identified as overpaid. A clear understanding of this difference in timeframes will help set the right expectations about how much $ can actually be recovered.

Question 37

Discuss the sample claim size in a typical claims audit

Answer 37

Almost no carrier/TPA will allow a full audit of 100% of the claims that were adjudicated during the allowed 12- to 24mo timeframe. Most auditors will apply a filtering process to the entire data file, in order to flag claims that show signs of having been paid incorrectly/over-/underpaid. From the filtered subset, the claims can then be chosen for the sample that would be a representative cross-section of the entire claim file.

If certain kinds of errors are found, this might indicate a systemic error, and the carrier/TPA would be instructed in the audit report to correct the problem going forward so that future claims are paid correctly.

Question 38

Discuss the types of samples that are used in administrative claims audit studies

Answer 38

A quality sample that is representative of the entire claim file can be generated only from a statistically valid random sample. Otherwise the quality and accuracy of the audit could be compromised. Audits conducted under the random sample method early present any opportunity for recoveries and are performed mainly for admin compliance purposes - that is, to ensure the carrier/TPA is administering the plan according to the plan doc.

The likelihood of the random sample including large-dollar claims with opps for recovery are slim, though not impossible. OTOH, if the ASO agrmt allows for it, the auditor can handpick the sample. This method allows for more of a focus on the large-dollar claims that may have been overpaid, and on where there is potential for a recovery from the provider.

A popular variation on these two methods is to split the sample: handpick the large-dollar claims (w/ recovery potential) using a portion of the sample, and randomly select the other portion in order to check for admin compliance with the plan doc.

Question 39

Outline the administrative claims audit process

Answer 39

Typically begin with a kickoff meeting/call where overview of process is presented to ensure sponsor’s objectives are fully defined & understood.

Auditor then collect info from sponsor & admin.

After receiving the data & info from both, most auditors will complete a comprehensive data analysis & scrubbing process. Auditor will need to verify key plan info/enrollment data to ensure an accurate audit.

The auditor then determines whether suspect claims were processed correctly.

Next, onsite portion of the audit: auditors have direct access to the admin’s system in order to validate potential over/underpayments and possible systemic errors/issues.

Once the onsite portion is complete, along with input from carrier/TPA, a draft audit report will be issued.

After the carrier provides feedback to the initial draft audit report, outstanding discrepancies will be addressed & final audit report issued.

When the final audit report has been reviewed, common practice is to schedule a call with all parties to discuss the process, results, outstanding issues, recommendations for improvements & the recovery process.

Question 40

What are 3 other types of audits that may be performed in conjunction with a primary claims audit?

Answer 40

1. Operational audits: look at procedures associated w enrollment, including ID card processing, customer service standards, etc.
2. Reinsurance audits: to discover charges that should have been submitted to a reinsurer for reimbursement but were not
3. Transitional audits: to ensure a plan has been set up appropriately in an administrator’s systems when a plan is moving to a new carrier, from fully ins to self-ins, or to significantly diff Ben designs

Question 41

Discuss the frequency of admin claims audits

Answer 41

Every 3-4y for most plan sponsors, if same carrier/TPA kept.

Some prefer more often.

Items that should trigger an audit:
1. A new carrier/admin
2. A significant plan design change
3. Admin concerns by the plan sponsor

Question 42

What is the typical cost of an audit?
Is this usually worth the investment?

Answer 42

Typically, <1% of total annual claim spending, and usually yes.

Most of the time, recoveries more than cover the cost, although that’s not guaranteed. Depending on size of the group & audit approach, total costs range ~$25k-50k, but could be higher for larger groups. Fees can be fixed or paid on contingency related to recoveries, although some ASO agreements don’t allow for contingency.