Analysis of POA and Hospital-Acquired Conditions Data: Part 2

by Lou Ann Schraffenberger, MBA, RHIA, CCS, CCS-P

Last October hospitals began reporting secondary diagnoses that were present on admission (POA). These indicators are reported as Y, present on admission; N, not present on admission; U, unknown; or W, clinically undetermined. Beginning October 1 of this year, N and U indicators will affect Medicare payment on eight hospital-acquired conditions selected by the Centers for Medicare and Medicaid Services.

As described in part 1 of this article, published in the June Journal, Advocate Health Care, an integrated healthcare delivery system of seven acute care hospitals in metropolitan Chicago, was interested in estimating the impact of this change on reimbursement. It analyzed data on its reporting of the POA indicator and the subsequent change in Medicare reimbursement in the fourth quarter of 2007.

This article reviews the 10 most frequent principal diagnoses reported with N or U for the non-Medicare population and the 10 most frequent secondary diagnoses reported with N or U for all patients. The first part of this article outlined the potential financial impact of the eight Medicare conditions on Advocate.

Most Frequent Principal Diagnoses

Using its data warehouse, Advocate produced frequency tables for principal and other diagnoses that were reported with POA indicator N or U during the fourth quarter of 2007. Lists of these diagnoses were collected for both Medicare payers and all other payers.

Advocate’s total number of discharges for the seven acute care hospitals for the fourth quarter 2007 totaled 35,500 patients. It discharged 13,100 Medicare beneficiaries from inpatient status for the same time period.

The 10 most frequent principal diagnoses reported with N or U for the non-Medicare population are shown in the table opposite. These principal diagnoses were reported a high of 299 times and a low of 19 times across Advocate’s seven hospitals.

Very few principal diagnoses were reported with N or U for the Medicare population. Alcohol withdrawal (291.81) was reported three times, and diabetes mellitus with circulatory complication with uncontrolled status was reported twice.

Other conditions that were reported once across the Medicare population included diabetes with renal manifestations, uncontrolled status; chronic duodenal ulcer with hemorrhage; unspecified septicemia; diverticulitis with hemorrhage; and asthma, unspecified, with acute exacerbation.

The conditions that were reported with the N or U status indicator appeared reasonable in terms of occurrence and the definition of principal diagnosis and POA guidelines.

Most of these codes reported with the N or U for the Medicare patients are combination codes; the patients had part of the condition when they were admitted, but another factor occurred during the hospital stay. For example, the diabetes became uncontrolled, the GI condition started to hemorrhage, or the asthma became exacerbated. With only a few of these patients reported (once across the Medicare population in this sample), it shows that this can occur, but not with any great frequency.

The 10 most frequent secondary diagnoses reported with N or U for all patients are shown in the table opposite. These other diagnoses were reported a high of 575 times to a low of 200 times across the Advocate’s seven hospitals:

Data Analysis and Reporting

Advocate shared the POA cumulative data and specific data for each hospital with the corporate-wide coding council. Diagnoses that were reported 50 or more times as N or U were examined for reasonableness; that is, whether the conditions could be acquired during the hospital stay. Conditions that attracted attention were dehydration, congestive heart failure, atrial fibrillation and atrial flutter, neonatal jaundice, and dysphagia.

The corporate-wide coding council includes the eight hospital coding managers and corporate support staff, including the HIM clinical data manager, internal auditor, and patient accounts and information systems liaisons who work to develop internal guidelines and best practices related to hospital coding and abstracting.

The manager of clinical data reviewed the same data with a physician vice president within the organization. The physician concluded it was possible that all the conditions may or may not be present on admission. However, both the coding leaders and the physician vice president noted that the four most frequent other diagnoses reported with N or U were conditions that could be diagnosed through laboratory tests.

Coders reported that physicians mentioned these diagnoses on day 2 or later during the hospital stay but not within the history, physical, or initial progress notes. Coders inquired whether they could review lab values for POA assignment. Often the lab values were abnormal on admission, but the physician did not mention the related condition until later in the stay.

Coders currently report an N or U when lab values on admission show abnormality but no related condition was documented as present on admission. On a number of occasions when the attending physician was queried about one of these three diagnoses (hypokalemia, anemia, UTI), the physicians responded with a yes or “look at the lab values.” When coders told the physicians they could not interpret lab values, the physicians responded, “Why not? The lab value is your answer.”

Most Frequent Principal Diagnoses Reported N or U, Non-Medicare Population

Shown here are the 10 most frequent principal diagnoses reported with N or U for Advocate’s non-Medicare population in fourth quarter 2007. These principal diagnoses were reported a high of 299 times and a low of 19 times across Advocate’s seven hospitals.

Rank

Principal Diagnosis

ICD-9-CM Diagnosis Code

1

First-degree perineal laceration with delivery

664.01

2

Second-degree perineal laceration with delivery

664.11

3

Abnormal fetal heart rate with delivery

659.71

4

Primary uterine inertia with delivery

661.01

5

Third-degree perineal laceration with delivery

664.21

6

Secondary uterine inertia with delivery

661.11

7

OB perineal trauma NEC with delivery

664.81

8

Complicated labor and delivery NEC with delivery

669.81

9

Other and unspecified uterine inertia with delivery

661.21

10

Precipitate labor with delivery

661.31


Most Frequent Secondary Diagnoses Reported N or U, All Patients

Shown here are the 10 most frequent secondary diagnoses reported with N or U for all Advocate patients in fourth quarter 2007. These other diagnoses were reported a high of 575 times to a low of 200 times across Advocate’s seven hospitals.

Rank

Secondary Diagnosis

ICD-9-CM Diagnosis Code

1

Hypopotassemia

276.8

2

Anemia, unspecified

285.9

3

Acute hemorrhagic anemia

285.1

4

Urinary tract infection, site not specified

599.0

5

Acute renal failure, unspecified

584.9

6

Acute respiratory failure

518.81

7

Pulmonary collapse/atelectasis

518.0

8

Pneumonia, organism unspecified

486

9

Hyposmolarity and/or hyponatremia

276.1

10

Constipation

564.00

Documentation, Reporting Improvement

Advocate will develop specific POA criteria to help physicians document certain conditions reported later in the stay (e.g., two or three days after the admission date) rather than rely upon POA lab values. Developing internal guidelines to provide coders with clinical guidance as to when a condition is present on admission will assist in accurate reporting; however, the guidance must be carefully applied to the POA indicator only and not to determine whether the condition is coded.

For example, a guideline indicating how lab values can be used to determine that hypokalemia was present on admission is acceptable, but the actual diagnosis of hypokalemia must be documented by the physician in order to be coded (e.g., on day 2 or 3 after admission). In other words, internal guidelines regarding lab values can be used to determine the POA indicator status but may not be used by coders to code a diagnosis the physician hasn’t documented.

Finally, Advocate will continue to work with physicians on documenting if a condition is present on admission. The organization will analyze data quarterly. Physicians are becoming more interested in how their documentation affects present on admission, hospital-acquired conditions, consumer access to provider information, and value-based purchasing.

References

Centers for Medicare and Medicaid Services (CMS). “Hospital-Acquired Conditions (Present on Admission Indicator).” Available online at www.cms.hhs.gov/HospitalAcqCond.

CMS. “Hospital-Acquired Conditions (HAC) in Acute Inpatient Prospective Payment System (IPPS) Hospitals.” December 2007. Available online at www.cms.hhs.gov/HospitalAcqCond/Downloads/hac_fact_sheet.pdf.

CMS. “Present on Admission (POA) Indicator Reporting by Acute Inpatient Prospective Payment System (IPPS) Hospitals.” December 2007. Available online at www.cms.hhs.gov/HospitalAcqCond/Downloads/poa_fact_sheet.pdf.

CMS. “Present on Admission Indicator.” MLN Matters, no. MM5499 (May 11, 2007). Available online at www.cms.hhs.gov/MLNMattersArticles/downloads/MM5499.pdf.

Deficit Reduction Act of 2005. Sec. 5001. Hospital Quality Improvement (c) Quality Adjustment in DRG Payments for Certain Hospital Acquired Infections. Available online at www.cms.hhs.gov/HospitalAcqCond/Downloads/DeficitReductionAct2005.pdf.

Illinois Hospital Association. “Illinois Consumer Guide Rules Adopted June 26, 2007.” Available online at www.ihatoday.org/issues/quality/consumerguide.html.

Illinois Hospital Association. “Present on Admission Reporting Requirements.” September 13, 2007. Available online at www.ihatoday.org/issues/quality/poa.html.

“Medicare: Changes to the Hospital Inpatient Prospective Payment Systems and Fiscal Year 2008 Rates; Final Rule.” Federal Register 72, no. 162 (August 22, 2007). Available online at www.access.gpo.gov/su_docs/fedreg/a070822c.html.

Lou Ann Schraffenberger (louann.schraffenberger@advocatehealth.com) is manager of clinical data in the Center for Health Information Services at Advocate Health Care in Oak Brook, IL.


Article citation:
Schraffenberger, Lou Ann. "Analysis of POA and Hospital-Acquired Conditions Data: Part 2" Journal of AHIMA 79, no.7 (July 2008): 70-72.