Legal and Regulatory Compliance

Compliance Analytics: There’s Gold in the Data

August 10, 2017 2:26 pm

Data analytics can help organizations determine risk areas for overpayment or underpayment, which ultimately improves the bottom line.

Healthcare finance and compliance professionals are rightly concerned about billing errors that result in overpayments, but they also should use data analytics—known as data mining—to find potential revenue opportunities from the unreported and underreported conditions that lead to underpayments. This was the essence of a presentation by Paul Belton, vice president for corporate compliance at Sharp HealthCare in San Diego, when he spoke at the Health Care Compliance Association’s regional conference recently in Orange County, California.

“The healthcare dollar is continually shrinking, so all payers are looking to maintain their surpluses,” Belton said. “The government needs to keep the Medicare Trust Fund solvent, Medicaid plans must combat state budget shortfalls, and private payers wish to maximize profits.” However, providers need to be paid properly for the care they give, and they can’t expect the government’s auditors to find underpayments.

Need to be Proactive

“Based on the tremendous disparity in external agencies’ audits and determinations of overpayments compared to underpayments, organizations need to take matters into their own hands in a proactive manner,” Belton said. “You must actively analyze your data to determine whether you are leaving money on the table due to missing modifiers, omitted procedure codes, incorrect DRG assignments, or various other causes.”

The emphasis should always be on clear, correct, and compliant documentation, coding, and DRG assignment because from a compliance perspective “underpayment is just as inappropriate as overpayment.” In addition to improving payment, data analytics can help determine how to identify the organization’s risk areas, protect resources, and improve clinical care for their patients.

Good data mining requires a team of prospectors that includes, at a minimum, coders, compliance auditors, clinical documentation improvement team members, and medical staff and revenue cycle representatives. The heads of those departments should also participate as necessary. The prospecting team should begin its work by running software—built in-house or purchased off the shelf—to identify claims likely to result in incorrect payment. Not every identified claim will contain errors, however, so expert auditors must conduct a secondary review of each flagged claim to ensure that proper DRGs and ICD-10 codes have been assigned.

Additional mining tools for benchmarking purposes include MedPAR data and PEPPER (Program for Evaluating Payment Patterns Electronic Report) Reports, the American Hospital Association’s RACTrac Survey Results, the Office of Inspector General (OIG) Workplan, and benchmarks of cohort facilities available from industry sources.

Searching for Nuggets

According to Belton, data mining is akin to prospecting for precious metal. “You’re looking for silver nuggets like unreported complications or comorbid conditions (CCs) and gold nuggets like major complications or comorbid conditions (MCCs).” He added that in addition to improving payment, you would likely uncover areas where you could reduce the risk of potential incorrect claims. Those findings might be even more valuable than the missed CCs and MCCs in terms of risk avoidance and financial loss.

Tool: Top 10 Claims Most Likely to be Flagged by Data Mining Software

Of course, with any mining operation there is always the potential to incur risk as well as reward.

For examples of conditions that have both, Belton cited the diagnoses of pneumonia and septicemia/sepsis requiring mechanical ventilation for more than 96 hours (DRGs 207/208 and 870/871/872). Claims for patients requiring lengthy ventilation are paid at a considerably higher rate than those for which the patient received less than 96 hours of ventilation, and they have therefore been the target of numerous government audits in recent years.

The OIG reported in 2013 and again in 2016 that Medicare has repeatedly paid hospitals improperly for these diagnoses, in some cases even paying the higher DRG when the patient’s entire length of stay (LOS) was less than the required 96 hours for ventilation alone. Being overpaid for such claims may subject the hospital to demands for repayment but also to the possibility of fraud allegations under the False Claims Act.  

On the flip side, data mining may reveal that poor documentation of mechanical ventilation start and end times has resulted in claims being underpaid. “As a prospector, you should be digging within the medical record for the silver and gold nuggets of those kinds of under-documented DRGs,” Belton stated.

Revenue Opportunities

According to Belton, some of the biggest revenue opportunities may come from the following areas:

  • DRGs with actual LOS significantly greater than the geometric mean LOS (GMLOS)
  • Claims with organ failure or with infectious disease as the principal diagnosis
  • Claims with sepsis as a secondary diagnosis with actual LOS greater than GMLOS
  • Cases with a missed cardiovascular CC, such as persistent atrial fibrillation, unstable angina, or atrial flutter
  • Mortality (expired patient) without an MCC
  • Medical or surgical cases without a CC or MCC

“In addition, I recommend digging deep whenever there are principal diagnoses like simple pneumonia, and digging even deeper when claims without CC or MCCs have secondary diagnoses such as hypotension, unspecified shock, unspecified congestive heart failure, and bedridden immobility,” Belton said.

To perform comprehensive clinical documentation and coding chart reviews, Belton uses the expertise and coding skills of Janice Amon Nozawa, MD, his organization’s corporate compliance clinical audit specialist. Nozawa explained, “We look for more details and review for additional clinical indicators, while having our CDI team consistently query and double check with the physicians. This usually results in more accurate coding and DRG assignment.

An education process begins once data analysis reveals patterns of improper documentation and/or coding. Education not only minimizes risk but also improves the bottom line through more accurate coding at the same time.”

Physician Education is Key

When asked about how best to get physicians to improve their documentation and thus billing accuracy, Belton offered three basic suggestions:

  • Meet with them one-on-one
  • Present evidence of how poor documentation affects payment
  • Have trusted and respected medical staff peers provide the training

“As part of this education process, the medical staff need to be encouraged to use recognized ICD-10 system terminology,” Belton said. “We show them where there are differences between clinical terms and the ICD-10 definitions and terminology, and we constantly remind them that the CDI team is there to help.”

“Compliance can very well become an asset,” Belton said, “and with data mining we have the opportunity to improve vague or incomplete clinical documentation, weak internal communication, overly conservative or inconsistent coding, and physician education—all of which should enhance the hospital’s bottom line and quality profile.”

“To be fully compliant, healthcare systems need to ensure that they capture all the revenue they’re entitled to and not a cent more,” Belton concluded. Thorough data mining and analysis will help achieve this optimum result.


J. Stuart Showalter, JD, MFS, is a contributing editor for HFMA.

Interviewed for this article:

Paul Belton, RHIA, MHA, MBA, JD, LLM, is vice president, corporate compliance, Sharp HealthCare, San Diego.

Janice Amon Nozawa, MD, CCDS, CCS, corporate compliance clinical audit specialist, Sharp Healthcare, San Diego.


googletag.cmd.push( function () { googletag.display( 'hfma-gpt-text1' ); } );
googletag.cmd.push( function () { googletag.display( 'hfma-gpt-text2' ); } );
googletag.cmd.push( function () { googletag.display( 'hfma-gpt-text3' ); } );
googletag.cmd.push( function () { googletag.display( 'hfma-gpt-text4' ); } );
googletag.cmd.push( function () { googletag.display( 'hfma-gpt-text5' ); } );
googletag.cmd.push( function () { googletag.display( 'hfma-gpt-text6' ); } );
googletag.cmd.push( function () { googletag.display( 'hfma-gpt-text7' ); } );
googletag.cmd.push( function () { googletag.display( 'hfma-gpt-leaderboard' ); } );