Final HIT Incentive Rule Quality Measures: Stroke Measure Set

Final HIT Incentive Rule Quality Measures: Stroke Measure Set

(Link to full report)

Developed by Beth Dibbert and Louis Wenzlow

This is the second in a series of 3 reports describing the final HIT incentive rule’s quality measure requirements. The first report, which summarizes the QI measure language in the final rule and focuses on the ER Throughput measure set, can be accessed at: http://www.worh.org/hit/2010/08/final-hit-incentive-rule-stage-1-quality-metrics-objective/. This report focuses on the Stroke measure set. Readers are encouraged to submit questions or offer additional information through the comment function.

Stroke Measure Set

The stroke process measures describe clinical decision-making and care for two adult inpatient populations: ischemic (caused by thrombosis or embolism) stroke or hemorrhagic (rupture of a blood vessel with bleeding into the tissue of the brain) stroke. Some measures apply to both populations, while others focus specifically on the care process for ischemic stroke patients. The identification of the stroke initial patient population for all of the Stroke Measures will include the following data elements:

• Principal Diagnosis Codes for ischemic or hemorrhagic stroke, (which is derived from the specific value set within the diagnosis codes having an active problem status and first priority)
• Patient Age (which is derived from a calculation of Admission Date minus Birthdate. Patients under the age of 18 on admission will be excluded)
• Length of Stay (which is a calculation derived from the Discharge Date minus Admission Date. Patients with a length of stay greater than 120 days are excluded from these measures)

1. Measure Stroke-2: Ischemic stroke – Discharge on anti-thrombotics (Ischemic stroke patients prescribed anti-thrombotic therapy at discharge: NQF #0435).

In addition to the data elements that identify the stroke initial patient population, the following data elements will be used to exclude patients to create a smaller denominator for this measure:

• Patients with Comfort Measures Only documented (Data elements required: procedure performed, problem code, diagnosis code, problem status)
• Patients enrolled in a Clinical Trial (Data elements required: problem code, diagnoses, problem status)
• Patients admitted for Elective Carotid Intervention (Data elements required: procedure performed, admission type)
• Patients discharged/transferred to another hospital for inpatient care (Data elements required: discharge disposition)
• Patients that left against medical advice, or discontinued care (Data elements required: discharge disposition)
• Patients expired (Data elements required: discharge disposition, time of death)
• Patients discharged/transferred to a federal healthcare facility (Data elements required: discharge disposition)
• Patients discharged/transferred to hospice (Data elements required: discharge disposition)
• Patients with a documented reason for not prescribing antithrombolytic therapy at discharge (Data elements required: procedure performed, problem code, diagnoses, discharge medication, tense, reason)

The numerator includes only those patients in the denominator population who were prescribed antithrombotic therapy at hospital discharge. Numerator capture will require that the patient’s discharge medications be compared to a value set of anti-thrombotic medications to determine whether anti-thrombotic therapy was prescribed.

Challenges:

1. As with the ED Throughput measures, some of the required data elements may or may not be discretely located within the electronic medical record (“Reason patients were not prescribed antithrombolytic therapy at discharge,” for instance). Presumably certified EHR vendors will be required to incorporate the appropriate discrete fields and make the required calculations. However, there may be significant workflow issues involved in entering the required data.
2. As with the ED Throughput measures, if the medical record is “missing” data element dates or times, the record will be rejected and not able to be calculated successfully. If the record is “unable to determine” data element dates or times, the measure will not be calculable. Both of these scenarios will require a collaborative “drill-down” by the facility and the vendor to determine whether the missing or undeterminable data is a result of workflow, technical, or a combination of causes.
3. As with the ED Throughput measures, the updates to data definitions have historically been a source of constant vigilance for vendors and hospitals alike. Care will need to be taken that the vendor applies the updates appropriately and in a timely manner. Updates to value sets, including medications value sets, will also need to be considered.
4. As with the ED Throughput measures, without appropriate vendor tools/reports that optimize the use of the data for internal purposes, hospitals personnel will find it difficult to apply quality improvement strategies and interventions to improve outcomes.
5. QI Abstractors are used to finding these elements in the “suggested data sources” listed in the national specifications manual. How certified EHRs will capture the many judgments, inferences and nuances that abstractors have been trained to interpret remains an open question.

2. Measure Stroke-3: Ischemic stroke – Anticoagulation for A-fib/flutter (Ischemic stroke patients with atrial fibrillation/flutter who are prescribed anti-coagulation therapy at hospital discharge: NQF #0436)

In addition to the data elements that identify the stroke initial patient population, the following data elements will be used to exclude patients to create a smaller denominator for this measure:

• Patients with Comfort Measures Only documented (Data elements required: procedure performed, problem code, diagnosis code, problem status)
• Patients enrolled in a Clinical Trial (Data elements required: problem code, diagnoses, problem status)
• Patients admitted for Elective Carotid Intervention (Data elements required: procedure performed, admission type)
• Patients discharged/transferred to another hospital for inpatient care (Data elements required: discharge disposition)
• Patients that left against medical advice, or discontinued care (Data elements required: discharge disposition)
• Patients expired (Data elements required: discharge disposition, time of death)
• Patients discharged/transferred to a federal healthcare facility (Data elements required: discharge disposition)
• Patients discharged/transferred to hospice (Data elements required: discharge disposition)
• Patients with a documented reason for not prescribing anticoagulation therapy at discharge (Data elements required: procedure performed, problem code, diagnoses, discharge medication, tense, reason)

The denominator population will now also include those ischemic stroke patients who have a diagnosis of Atrial Fibrillation/Flutter (Data elements required: problem code, diagnoses, procedure performed)

The numerator includes only those patients in the denominator population who were prescribed anticoagulation therapy at hospital discharge. Numerator capture will require that the patient’s discharge medication information be compared to a value set of anticoagulation medications to determine whether anticoagulation therapy was prescribed.

Challenges:

1. The same challenges associated with Stroke-2 apply.
2. Abstracting the “documented reason for not prescribing anticoagulation therapy” will likely be the greatest data collection challenge.

3. Measure Stroke-4: Ischemic stroke – Thrombolytic therapy for patients arriving within 2 hours of symptom onset (ischemic stroke patients who arrive at this hospital within 2 hours of the time last known well and for whom IV t-PA was initiated at this hospital within 3 hours of the time last known well: NQF #0437).

In addition to the data elements that identify the stroke initial patient population, the following data elements will be used to exclude patients to create a smaller denominator for this measure:

• Patients with Comfort Measures Only documented (Data elements required: procedure performed, problem code, diagnosis code, problem status)
• Patients enrolled in a Clinical Trial (Data elements required: problem code, diagnoses, problem status)
• Patients admitted for Elective Carotid Intervention (Data elements required: procedure performed, admission type)
• Time last known well to arrival in the emergency department greater than 2 hours (problem date/time, diagnosis date/time, arrival date/time, problem code, discharge diagnosis code)
• Patients with documented reason for not initiating IV Thrombolytic (Data elements required: procedure performed, problem code, diagnoses, medications administered, tense, reason)

The numerator includes only those patients in the denominator population who were given IV thrombolytic therapy within 3 hours (180 minutes) of time last known well. Numerator capture will require that the patient’s medication information (medication administered date/time, medications administered, medications administered route) be compared to a value set of thrombolytic medications to determine whether thrombolytic therapy was administered within the 3 hours.

1. The same challenges associated with Stroke-2 apply.
2. Calculating this measure involves greater complexity, due to the two “date/time last known well” calculations, and the requirement for additional medication information.
3. Establishing “date/time last known well” requires protocols that solicit the best possible information from EMS, family members, and others. This is a challenge whether the data needs to be entered into the certified EHR or not.
4. Abstracting the “documented reason for not initiating IV Thrombolytic” will likely be the greatest data collection challenge.

4. Stroke-5: Ischemic stroke – Antithrombotic therapy by day 2 (Ischemic stroke patients administered antithrombotic therapy by the end of hospital day 2: NQF #0438).

In addition to the data elements that identify the stroke initial patient population, the following data elements will be used to exclude patients to create a smaller denominator for this measure:

• Patients discharged by end of hospital day 2 (Data elements required: admission date/time, discharge date/time)
• Patients with Comfort Measures Only documented on day of or day after arrival (Data elements required: procedure performed, problem code, diagnosis code, problem status, procedure date, admit date/time)
• Patients enrolled in a Clinical Trial (Data elements required: problem code, diagnoses, problem status)
• Patients admitted for Elective Carotid Intervention (Data elements required: procedure performed, admission type)
• Patients with IV or IA Thrombolytic (t-PA) Therapy administered at this hospital or within 24 hours prior to arrival (Data elements required: arrival date/time, medications administered, medications administered route, medications administered date/time)
• Patients with documented reason for not initiating antithrombolytic therapy by end of hospital day 2 (Data elements required: procedure performed, problem code, diagnoses, discharge medication, tense, reason)

The numerator includes only those patients in the denominator population who were given antithrombolytic therapy by end of hospital day 2. Numerator capture will require that the patient’s medication information (medication administered date/time, medications administered, admit date/time) be compared to a value set of antithrombolytic medications to determine whether antithrombolytic therapy was administered by day 2.

Challenges:

1. The same challenges associated with Stroke-2 apply.
2. With this measure there is an additional calculation for “duration of stay,” and Comfort Measures Only must be linked to an order date.
3. Abstracting the “documented reason for not prescribing antithrombolytic therapy” will likely be the greatest data collection challenge.
4. The final rule includes “hemorrhagic stroke” patients as being included in this measure, however, only ischemic stroke patients are listed in the description. We believe that this is a typographical error in the final rule measure title, and that the measure applies to ischemic stroke patients only.

5. Stroke-6: Ischemic stroke – Discharge on statins (Ischemic stroke patients with LDL greater than or equal to 100mg/dL, or LDL not measured, or, who were on a lipid-lowering medication prior to hospital arrival are prescribed statin medication at hospital discharge (NQF #0439).

In addition to the data elements that identify the stroke initial patient population, the following data elements will be used to exclude patients to create a smaller denominator for this measure:

• Patients with pre-arrival lipid-lowering agent (Data elements required: arrival date/time, medications administered, )
• Patients with LDL-c not measured[1] (Data elements required: arrival date/time, result type, report date/time)
• Patients with LDL-c greater than or equal to 100 mg/dL (Data elements required: result type, result value)

• Patients with Comfort Measures Only documented (Data elements required: procedure performed, problem code, diagnosis code, problem status)
• Patients enrolled in a Clinical Trial (Data elements required: problem code, diagnoses, problem status)
• Patients admitted for Elective Carotid Intervention (Data elements required: procedure performed, admission type)
• Patients without evidence of Atherosclerosis[2] (Data elements required: problem code, diagnoses)
• Patients discharged/transferred to another hospital for inpatient care (Data elements required: discharge disposition)
• Patients who left against medical advice or discontinued care (Data elements required: discharge disposition)
• Patient expired (Data elements required: discharge disposition, time of death)
• Discharged/transferred to a federal healthcare facility (Data elements required: discharge disposition)
• Discharged/transferred to hospice (Data elements required: discharge disposition)
• Patients with a documented reason for not prescribing statin medication at discharge (Data elements required: problem code, diagnoses, discharge medication, tense, reason)

The numerator includes only those patients in the denominator population who were prescribed statin medication at discharge. Numerator capture will require that the patient’s medication information (discharge medications orders) be compared to a value set of statin medications.

Challenges:

1. The same challenges associated with Stroke-2 apply.
2. Abstracting the pre-arrival statin use information, as well as the “documented reason for not prescribing statin medication at discharge,” will likely be the greatest data collection challenges.

6. Stroke-8: Ischemic or hemorrhagic stroke – Stroke education (Ischemic or hemorrhagic stroke patients or their caregivers who were given educational materials during the hospital stay addressing all of the following: (1) Activation of EMS, (2) Need for follow-up after discharge, (3) Medications prescribed at discharge, (4) Risk factors for stroke, (5) Warning signs and symptoms of stroke NQF #0440).

Ischemic stroke patients as well as hemorrhagic stroke patients make up the denominator population. In addition to the data elements that identify the stroke initial patient population, the following data elements will be used to exclude patients to create a smaller denominator for this measure:

• Patients with Comfort Measures Only documented (Data elements required: procedure performed, problem code, diagnosis code, problem status)
• Patients enrolled in a Clinical Trial (Data elements required: problem code, diagnoses, problem status)
• Patients admitted for Elective Carotid Intervention (Data elements required: procedure performed, admission type)
• Patients not discharged to home or homecare (Data elements required: discharge disposition)

The numerator includes only those patients in the denominator with documentation that they or their caregivers were given educational material addressing all of the critical elements listed in the measure description above. Numerator capture will require that procedures performed be compared to value sets for documentation of findings/procedures.

Challenges:

1. The same challenges associated with Stroke-2 apply.
2. This is an “all-or-nothing” measure, which means each element of education (EMS activation AND need for follow-up AND medications, etc….) must be present in order for the patient to be included in the numerator. Currently the core measures specifications allow for abstractors to use situational inference and exclusions that are not found in HITSP specifications. For instance, if the patient was not given EMS activation information because they have a cognitive impairment and has no caregiver, the abstractor is allowed to select “yes” on an abstraction form, as if the patient had received the educational material. How will such decision making be accommodated by automated programming?

7. Stroke-10: Ischemic or hemorrhagic stroke – Rehabilitation assessment (Ischemic or hemorrhagic stroke patients or their caregivers who were assessed for rehabilitation services: NQF #0441).

Ischemic stroke patients as well as hemorrhagic stroke patients make up the denominator population. In addition to the data elements that identify the stroke initial patient population, the following data elements will be used to exclude patients to create a smaller denominator for this measure:

• Patients with Comfort Measures Only documented (Data elements required: procedure performed, problem code, diagnosis code, problem status)
• Patients enrolled in a Clinical Trial (Data elements required: problem code, diagnoses, problem status)
• Patients admitted for Elective Carotid Intervention (Data elements required: procedure performed, admission type)
• Patients discharged/transferred to another hospital for inpatient care (Data elements required: discharge disposition)
• Patients that left against medical advice or discontinued care (Data elements required: discharge disposition)
• Patients expired (Data elements required: discharge disposition, time of death)
• Patients discharged/transferred to a federal healthcare facility (Data elements required: discharge disposition)
• Patients discharged/transferred to hospice (Data elements required: discharge disposition)

The numerator includes only those patients in the denominator who received an assessment for rehabilitation services. Data elements required for numerator capture will include problem code, diagnoses, and procedure performed, all of which will need to be compared to “documentation of findings” value sets.

Challenges:

1. The same challenges associated with Stroke-2 apply.
2. As with Stroke-8, abstractor guidance now allows a “yes” response (indicating that an assessment was done or rehab services took place) in the event documentation exists for not completing an assessment. Examples of such documentation include patient/family refusal, notes that the patient was unable to tolerate rehab therapy, etc. Will MU automation disallow such exceptions or will abstractors be able to use guidance from the “manual version” to override programming?

---

[1] LDL-c data element can be derived from information as far back as 30 days prior to hospital arrival. How do hospitals obtain this data now, and how will implementation towards meaningful use incorporate and modify this workflow?

[2] We are listing these specifications as they are written in HITSP, which is referenced on the CMS website. Note that the specifications are already being revised for October 2010, and that this data element does not exist for the purposes of calculating this measure.