Mortality Measurement: Development of the 3M™ All Patient Refined Diag

Presentations from a November 2008 meeting to discuss issues related to mortality measures.

Determination of the Risk of Mortality Subclass

The three-phase process of determining the risk of mortality subclass is summarized in Figure 2-2. This three-phase process parallels the three phases in the determination of the severity of illness subclass. In Phase I, the risk of mortality of each secondary diagnosis is determined. Once the risk of mortality level of each individual secondary diagnosis is established, then Phase II determines a base risk of mortality subclass for the patient based on all of the patient's secondary diagnoses. In Phase III, the final subclass for the patient is determined by incorporating the impact of principal diagnosis, age, OR procedures, certain non-OR procedures, multiple OR procedures, and combinations of categories of secondary diagnoses.

 

Figure 2-2. Three-phase process for determining patient risk of mortality subclass

Figure depicts the three-phase process of determining the risk of mortality subclass. For details, go to [D] Text Description.

[D] Select for Text Description.

Phase I—Determining the risk of mortality level of each secondary diagnosis

1. Eliminate secondary diagnoses associated with the principal diagnosis

This step is identical to the corresponding step in the determination of the severity of illness subclass. If a secondary diagnosis is closely related to the principal diagnosis and does not add any distinguishing information, then the secondary diagnosis is completely excluded from the 18 step process to determine the patient's risk of mortality subclass.

2. Assign each secondary diagnosis its standard risk of mortality level

Each secondary diagnosis is assigned one of four distinct risk of mortality levels. In general, except for malignancies and certain extreme acute diseases such as acute renal failure, the risk of mortality level tends to be lower than the severity of illness level for the same diagnosis. Mortality is relatively rare. There are a limited number of diagnoses that significantly increase the risk of mortality. For example, traumatic amputation of the arm, acute cholecystitis, and acute osteomyelitis are all at a major severity of illness level since they represent serious diseases with significant loss of organ function. However, they present relatively low risk of mortality and therefore are assigned to a minor risk of mortality level. Example of secondary diagnoses that would have an extreme risk of mortality are intracranial hemorrhage, acute vascular insufficiency of intestine, acute myocardial infarct, and acute renal failure.

For version 20.0 APR DRGs, the standard risk of mortality level was comprehensively reviewed for all secondary diagnoses codes. There were a number of revisions introduced, the majority of which were to lower the standard risk of mortality level. In situations where there was a great deal of variability within an ICD-9-CM diagnosis code, the approach was to lower the standard risk of mortality level and then in later steps of Phase I, consider whether modifications to the standard risk of mortality level are indicated based upon specific age ranges, APR DRGs, or non-OR procedures.

For version 25.0 APR DRGs, there are a total of 13,677 ICD-9-CM diagnosis codes. These codes are assigned to the following risk of mortality levels: 10,996 minor, 1,713 moderate, 617 major, 351 extreme. For version 25.0 APR DRGs, there are 2,681 secondary diagnosis codes that are assigned a standard risk of mortality level of moderate, major, or extreme. This is just slightly more than half the 4,884 secondary diagnosis codes that are assigned a standard severity of illness level of moderate, major, or extreme.

3. Modify the standard risk of mortality level of a secondary diagnosis based on age

The standard risk of mortality for certain secondary diagnoses may be modified depending upon the age of the patient. This age modification is applied much more extensively for risk of mortality, than for severity of illness. For pediatric patients, the standard risk of mortality level of secondary diagnoses is often decreased. For example, the risk of mortality level for diabetes with ketoacidosis is lowered from moderate to minor for pediatric patients. It is also lowered for many other secondary diagnoses including infectious illnesses and traumatic injuries. However, for some pediatric diagnoses, mostly congenital anomalies, the risk of mortality level is increased during the neonatal time period and sometimes the first year of life. For example, the risk of mortality level for hypoplastic left heart syndrome is increased from major to extreme during the neonatal period; renal dysphasia is increased from moderate to major during the neonatal period; and congenital tricuspid atresia/stenosis is increased from moderate to major during the first year of life.

For elderly patients, the standard risk of mortality level is increased to a higher level for many secondary diagnoses. Elderly patients are most often defined as age >65 years or age >69 years but also sometimes for a more narrowly defined subset of elderly patients such as age >79 years. For example, for elderly patients age >65 years the risk of mortality level is increased from minor to moderate for secondary diagnoses such as atrial fibrillation, chronic obstructive lung disease and nephritis, and is increased from moderate to major for acidosis and hypotension. For elderly patients age >69 years, the risk of mortality level is increased from minor to moderate viral pneumonia, mitral valve disorder, and anemia; and from moderate to major for streptococcal, staphylococcal, and other bacterial pneumonias; and from major to extreme for peritonitis. For elderly patients age >79 years, the risk of mortality level is increased from minor to moderate for fracture of femur or pelvis; and from moderate to major for pleural effusion.

4. Modify the standard risk of mortality level of a secondary diagnosis based on the APR DRG and principal diagnosis

The standard risk of mortality level for some secondary diagnoses may be modified depending on the APR DRG and principal diagnosis of the patient. In versions 20.0 and 25.0, this logic is applied only to APR DRG 190 Acute Myocardial Infarct. In general, secondary diagnoses that are closely related to the principal diagnosis are excluded from the determination of the risk of mortality subclass. However, for a patient admitted for an acute anterior wall myocardial infarction, an acute anterolateral myocardial infarction represents an extension of the acute anterior wall myocardial infarction. Therefore, the acute anterolateral myocardial infarction is not excluded and is assigned a risk of mortality level of moderate.

5. Modify the standard risk of mortality of a secondary diagnosis based on the APR DRG

The standard risk of mortality level for many secondary diagnoses is modified depending upon the APR DRG to which the patient is assigned. Altogether, there are 1,474 modifications of the standard risk of mortality level of secondary diagnosis depending on the APR DRG. As with severity of illness, the APR DRG specific modifications to the risk of mortality level of individual secondary diagnoses reflects the disease-specific nature of the determination of risk of mortality.

For example, the risk of mortality level for secondary diagnoses is increased from minor to moderate for the following combinations of secondary diagnoses and APR DRGs: right bundle branch block and APR DRG for acute myocardial infarct; chronic obstructive lung disease and major chest and major cardiovascular surgery; hypovolemia and APR DRGs for cancer, cardiovascular disease, and respiratory failure. The risk of mortality level for secondary diagnoses is increased from moderate to major for the following combinations of secondary diagnoses and APR DRGs: acidosis and APR DRGs for acute myocardial infarct, congestive heart failure, and septicemia; hypotension and APR DRGs for respiratory failure, acute myocardial infarct, and liver and pancreas disorders.

There are also many APR DRGs where the standard risk of mortality level for some secondary diagnoses is decreased, such as for secondary diagnoses that are closely related to the definition of the APR DRG. For example, the risk of mortality level is decreased from moderate to minor for secondary diagnosis of obstructive hydrocephalus in the APR DRG for ventricular shunt procedures, since the hydrocephalus is the underlying reason for performing the procedure. The risk of mortality level is decreased from extreme to major for secondary diagnosis of cerebral edema in a number of nervous system APR DRGs including craniotomy, cerebrovascular disease, and malignancy. If there is essentially complete overlap between the secondary diagnosis and the APR DRG, the risk of mortality level for the secondary diagnosis may be decreased from extreme or major to minor. For example, acute respiratory failure is decreased from extreme to minor for APR DRGs for respiratory system diagnosis with mechanical ventilation 96+ hours and tracheostomy with mechanical ventilation 96+ hours. There are many secondary diagnoses for which the standard risk of mortality level is lowered to minor for a patient in one of eleven elective, non-extensive surgical APR DRGs. For example, in these APR DRGs, secondary diagnoses of malignant neoplasm are reduced from major or moderate to minor, since the patient would likely not have these surgical procedures performed if the malignancy was at a stage that represented a significant risk of mortality.

6. Modify the standard risk of mortality level of a secondary diagnosis based on non-OR procedure

Certain non-OR procedures will sometimes be used to modify the standard risk of mortality level of some secondary diagnoses. For risk of mortality, this step is just used with one non-OR procedure, pulsation balloon implant. For example, subendocardial infarction has a standard risk of mortality level of moderate but is increased by an increment of two up to extreme if the patient had a pulsation balloon implanted. The need for the pulsation balloon is an indicator of the extent of the subendocardial infarction.

Phase II—Determine the base risk of mortality subclass for the patient

Once each secondary diagnosis has been assigned its standard risk of mortality level and the standard risk of mortality level of each secondary diagnosis has been modified based on the patient's age, APR DRG and principal diagnosis, APR DRG, and certain non-OR procedure, the Phase II base risk of mortality subclass for the patient can be determined. The process of determining the base patient risk of mortality subclass begins with the elimination of certain secondary diagnoses that are closely related to other secondary diagnoses. The elimination of these diagnoses prevents the double counting of clinically similar diagnoses in the determination of the risk of mortality subclass of the patient. Once redundant diagnoses have been eliminated, the base risk of mortality subclass is determined based on all of the remaining secondary diagnoses. There are three steps to Phase II for risk of mortality. The first two are the same as for severity of illness. The third step is similar to severity of illness but has some additional exceptions logic.

7. Eliminate certain secondary diagnoses from the determination of the risk of mortality subclass of the patient

This step is identical to the corresponding step in the determination of the severity of illness subclass. Secondary diagnoses that are related to other secondary diagnoses have their risk of mortality level reduced to minor.

8. Combine all secondary diagnoses to determine the base risk of mortality subclass of the patient

Once secondary diagnoses that are related to other secondary diagnoses have their risk of mortality level reduced to minor, the base patient risk of mortality subclass is set equal to the maximum risk of mortality level across all of the remaining secondary diagnoses. This is done the same way as for severity of illness. For example, if there are five remaining secondary diagnoses and one is a major risk of mortality level and four are a moderate risk of mortality level, then the base patient risk of mortality subclass is major.

9. Reduce the base risk of mortality subclass if the patient does not have multiple secondary diagnoses with a significant risk of mortality, except for certain secondary diagnoses for which this requirement is removed or modified

In general, high risk of mortality patients are characterized by multiple secondary diagnoses with a significant risk of mortality. In order for the base risk of mortality subclass to be extreme, there must be two or more extreme risk of mortality secondary diagnoses present or a single extreme risk of mortality secondary diagnosis plus two or more major risk of mortality secondary diagnoses. If this multiple criteria is not met, the patient's base risk of mortality subclass is lowered to either major or moderate. If the multiple criteria is not met, but in addition to a single extreme risk of mortality secondary diagnosis there is at least one other major or moderate secondary diagnosis, then the patient's risk of mortality subclass is lowered to major. If there is not at least one other major or moderate secondary diagnosis in addition to an extreme risk of mortality secondary diagnosis, then the patient's base risk of mortality subclass is lowered to moderate.

There are, however, two exceptions to these criteria. There is one set of secondary diagnoses that have such an inherent high risk of mortality that no other secondary diagnoses are required for the patient's base risk of mortality subclass to be extreme. Examples include: pulmonary anthrax, ruptured aortic aneurism, hepatorenal syndrome, head trauma with deep coma, and 60-90% body burn/50-59% third degree. There is a second set of secondary diagnoses that also have an inherently high risk of mortality and for which only one other major secondary diagnosis is required for the patient's base risk of mortality to be extreme. Examples included: defibrination syndrome, acute myocardial infarct, intracranial hemorrhage, cerebral thrombosis with infarct, dissection of aortic aneurism, acute respiratory failure, acute renal failure, and shock.

Patients with a base risk of mortality subclass of major are reduced to moderate unless, in addition to the major risk of mortality secondary diagnosis, there is at least one additional major risk of mortality secondary diagnosis or two more additional secondary diagnoses with a moderate risk of mortality. If this multiple criteria is not met then the patient's base risk of mortality subclass is lowered to moderate. There are, however, two exceptions to these criteria. There is one set of secondary diagnoses that have a sufficiently high inherent risk of mortality that no other secondary diagnoses are required for the patient's base risk of mortality subclass to be set at major. Examples include: flail chest, major liver laceration, 40-49% body burns/10-19% third degree. There is a second set of secondary diagnoses that have a significant inherent risk of mortality so that only one moderate secondary diagnoses is required for the patient's base risk of mortality subclass to be set at major.

Examples include: food/vomit pneomonitis, acute lung edema, and perforation of intestine. Patients with a base risk of mortality subclass of moderate are reduced to minor unless there are at least two moderate risk of mortality secondary diagnoses present. There is, however, one exception to this criteria. These moderate risk of mortality secondary diagnoses do not require any other secondary diagnoses to be present. Examples include: malignant neoplasm diagnoses that are moderate risk of mortality level diagnoses, acidosis, bacterial pneumonia, congestive heart failure, chronic renal failure, Alzheimer's disease, and decubitus ulcer.

Phase III—Determine the final risk of mortality subclass of the patient

Once the base patient risk of mortality subclass is computed then the risk of mortality subclass may be increased or decreased in Phase III based on specific values of certain patient attributes. In Phase III, the risk of mortality algorithm examines six of the eight patient attributes utilized in Phase III of the severity of illness logic. The two that are not used by risk of mortality are only used to a very limited extent in the severity of illness logic. The patient attributes are:

  • Combinations of APR DRG and principal diagnosis
  • Combinations of APR DRG and age, or APR DRG and principal diagnosis and age, or APR DRG and birthweight and absence of certain non-OR procedures
  • Combinations of APR DRG and non-OR procedures
  • Combinations of APR DRG and OR procedures
  • Combinations of APR DRG and pairs of OR procedures
  • Combination of the APR DRG for ECMO and presence/absence of certain OR procedures (not applicable for risk of mortality)
  • Combinations of APR DRG and principal diagnoses and non-OR procedures (not applicable for risk of mortality)
  • Combinations of categories of secondary diagnoses

In Phase I, age and non-OR procedures were used to modify the standard risk of mortality level of a secondary diagnosis. However, age and non-OR procedures can also have an impact that is specific to the patient's APR DRG or a specific principal diagnosis within an APR DRG. Thus, the impact of age and non-OR procedures is reassessed as part of the determination of the risk of mortality subclass of the patient. Based on the patient attributes listed above, a series of modifications to the base patient risk of mortality subclass are made during Phase III. The final patient risk of mortality subclass will be computed based on the Phase II base patient risk of mortality subclass and the modifications to the base risk of mortality subclass made in Phase III.

10. Modify the risk of mortality subclass for the patient based on the APR DRG and principal diagnosis

Within specific APR DRGs some principal diagnoses are indicative of higher or lower risk of mortality relative to the other principal diagnoses in the APR DRGs. This is one of the most important and extensively used modifications to the patient's base risk of mortality subclass that occurs as part of the Phase III risk of mortality logic. The majority of the modifications are increases to the patient risk of mortality subclass, but there are also some decreases to the patient risk of mortality subclass. Some of the increases are an increment of one up to a maximum subclass of moderate, while others pertain to more dramatic clinical situations and provide greater increases to the patient risk of mortality subclass.

Most of the decreases reduce the patient risk of mortality subclass by one from major or moderate. Following are examples:

  • APR DRG 309 Hip & Femur Procedures For Non-Trauma Except Joint Replacement and principal diagnosis of secondary malignancy of bone: increase patient risk of mortality subclass by one up to a maximum of moderate.
  • APR DRG 135 Major Chest & Respiratory Trauma and principal diagnosis of flail chest: increase patient risk of mortality subclass by one up to a maximum of major.
  • APR DRG 221 Major Large & Small Bowel Procedures and principal diagnosis of perforation of intestine: increase patient risk of mortality subclass by two up to a maximum of major.
  • APR DRG 169 Major Thoracic & Abdominal Procedures and principal diagnosis of ruptured abdominal aortic aneurism: increase patient risk of mortality subclass by three up to extreme.
  • APR DRG 44 Intracranial Hemorrhage and principal diagnosis of subdural hemorrhage: decrease patient risk of mortality subclass by one from moderate.
  • APR DRG 52 Non traumatic Stupor & Coma and principal diagnosis of transient alteration of awareness: decrease patient risk of mortality subclass by one from extreme, major, or moderate.

11. Modify the risk of mortality subclass for the patient based on combinations of the APR DRG and principal diagnosis and age, or APR DRG and age, or APR DRG and birthweight and presence/absence of certain non-OR procedures

For some principal diagnoses in specific APR DRGs, the patient's age essentially represents a complicating factor. For specific principal diagnosis and age combinations in certain APR DRGs, the risk of mortality subclass of the patient is increased by a specified increment up to a specified maximum subclass. For example, elderly patients age >79 years in APR DRG 137 Major Respiratory Infections & Inflammations with a principal diagnosis of staphylococcal pneumonia and elderly patients age >79 years in APR DRG 710 Septicemia & Disseminated Infections with most of the septicemia principal diagnoses, have their risk of mortality subclass increased by one up to a maximum subclass of moderate. Elderly patients age >69 years in APR DRG 44 Intracranial Hemorrhage with a principal diagnosis of intracerebral hemorrhage have their risk of mortality subclass increased by one up to a maximum subclass of moderate. The increase indicates that intracranial hemorrhage in an elderly patient represents a higher risk of mortality.

This step is also sometimes implemented for all patients in a specified age range in an APR DRG rather than just for patients with a particular principal diagnoses. This approach is used for elderly patients age >84 years for 19 APR DRGs involving major surgery. For example, patients age >84 years in APR DRG 120 Major Chest & Respiratory Procedures have their risk of mortality subclass increased by one to a maximum subclass of moderate.

The last part of this step examines the relationship between APR DRG and birthweight and presence/absence of certain non-OR procedures for extremely low birthweight neonates in MDC 15. Many of the neonates at an extremely low birthweight (<750 grams or 1.6 pounds) are non-viable and receive comfort-only care. Nearly all of these newborns die and most of the time this is within a few days of being born. There are no ICD-9-CM diagnosis codes for non-viability due to extreme prematurity, which, if such codes existed, would allow a risk of mortality subclass of extreme to be assigned. In its place, the APR DRG system has developed logic to identify these cases. Since newborns <750 grams will virtually always receive some therapeutic interventions if the goal is to maintain life (e.g., respiratory therapy, tube feedings), the absence of any of these non-OR procedures can be used to infer the newborn is receiving comfort-only measures and their risk of mortality subclass is increased to extreme for APR DRGs 589 and 591. Without this logic, most of these newborns would be a risk of mortality subclass minor or moderate because of the lack of codes for identifying non-viability.

12. Modify the risk of mortality subclass for the patient based on combinations of APR DRG and non-OR procedure

For some APR DRGs the presence of certain non-OR procedures is indicative of a more extensive disease process with a higher risk of mortality. In these instances, the risk of mortality subclass is increased by a specific increment up to a specified maximum. There are three non-OR procedures used for this step: mechanical ventilation 96+ hours, mechanical ventilation <96 hours, and balloon pulsation device. For example, for patients in APR DRG 194 Heart Failure the risk of mortality subclass is increased by two up to a maximum subclass of extreme if mechanical ventilation 96+ hours is performed and is increased by one up to a maximum subclass of major if mechanical ventilation <96 hours is performed.

13. Modify the risk of mortality subclass for the patient based on combinations of APR DRG and OR procedure

Within specific APR DRGs, some OR procedures are indicative of higher risk of mortality relative to the other OR procedures in the APR DRG. For example, the risk of mortality subclass of patients in APR DRG 443 Kidney and Urinary Tract Procedures for Non-Malignancy, is increased by two up to a maximum of major if the procedure bilateral nephrectomy is performed. Relative to other procedures in DRG 443, a bilateral nephrectomy represents a patient that has a higher risk of mortality.

Within specific APR DRGs, there are also some OR procedures that are indicative of lower risk of mortality relative to other patients in the same APR DRG. For example, a patient in APR DRG 220 Major Stomach Esophageal & Duodenal Procedures who receives a procedure to create esophogastric sphincteric competence has a lower risk of mortality than other surgical patients in APR DRG 220 (e.g., esophagectomy, gastrectomy), and if up to this point in the process their risk of mortality subclass is moderate, it is decreased by 1 to minor.

14. Modify the risk of mortality subclass for the patient based on combinations of APR DRG and pairs of OR procedures

Within specific APR DRGs the presence of certain pairs of OR procedures is indicative of a more extensive disease process and a higher risk of mortality relative to other patients in the same APR DRG. For risk of mortality, this logic is applicable primarily for patients who receive both a peripheral bypass procedure and a lower limb amputation. For example, a patient in either APR DRG 173 Other Vascular Procedures or APR DRG 305 Amputation of Lower Limb who receives both a peripheral bypass procedure and a lower leg amputation has their risk of mortality subclass increased by an increment of one up to a maximum subclass of major.

15. Modify the risk of mortality subclass for the patient based upon combination of the APR DRG for ECMO and presence/absence of certain OR procedures

This step is not applicable to risk of mortality.

16. Modify the patient risk of mortality subclass based on the APR DRG and principal diagnosis and certain non-OR procedures

This step is not applicable to risk of mortality.

17. Establish a minimum risk of mortality subclass for the patient based on combinations of categories of secondary diagnoses

The presence of certain combinations of secondary diagnoses has great clinical significance. The interaction of specific combinations of secondary diagnoses increases the risk of mortality. Therefore, a minimum patient risk of mortality subclass greater than subclass minor is established if certain combinations of secondary diagnoses are present. The presence of multiple interacting diagnoses is characteristic of high risk of mortality patients. A subset of secondary diagnoses will interact with each other causing patient risk of mortality to be increased.

The categories of secondary diagnoses used for this step in risk of mortality are the same 83 core secondary diagnosis categories that are used for severity of illness (see table 2-3). The only difference is that these same 83 secondary diagnosis categories are then subdivided by risk of mortality level, not severity of illness level. The additional 21 secondary diagnosis categories developed for use with neonatal APR DRGs 626 and 640 are not used for risk of mortality. These additional 21 secondary diagnosis categories are intended to differentiate neonates with multiple minor or other problems from those who are normal newborns or who have a single minor problem, which is significant for severity of illness but is not applicable for risk of mortality since these diagnoses do not increase the risk of dying.

All of the secondary diagnosis category combination types for risk of mortality are the same as those defined for severity of illness (Table 2-5). Of the nine possible combination types, six are applicable for risk of mortality. These are combination types 1, 2, 3, 4, 5, and 13.

A type 1 combination consists of two categories that contain major risk of mortality level diagnoses, plus any two additional secondary diagnoses that are at least major level. When a type 1 combination occurs, the minimum patient risk of mortality subclass is extreme. An example of a type 1 combination is a major pulmonary diagnosis (category 75) such as acute pulmonary edema and a major neurological diagnosis (category 64) such as cerebral thrombosis without infarct combined with any other two major secondary diagnoses.

A type 2 combination is the same as type 1 except that the two categories consist of a major risk of mortality category and a moderate risk of mortality category. For a type 2 combination, the minimum patient risk of mortality subclass is extreme. An example of a type 2 combination is a major bacterial infection (category 9) such as peritonitis and a moderate level secondary malignancy (category 78) combined with any other two major secondary diagnoses.

A type 3 combination consists of two categories that contain moderate risk of mortality level diagnoses, plus any two additional secondary diagnoses that are at least a moderate risk of mortality level. For a type 3 combination, the minimum patient risk of mortality is major. An example of a type 3 combination is a moderate bacterial infection (category 9) such as staphylococcal enteritis with chronic renal failure (category 20) combined with any other two moderate secondary diagnoses.

A type 4 combination consists of a moderate risk of mortality category and a minor risk of mortality category, plus any two additional secondary diagnoses that are at least moderate. For a type 4 combination, the minimum patient risk of mortality subclass is major. An example of a type 4 combination is a decubitus ulcer (category 26) and hypovolemia (category 51) combined with two other secondary diagnoses that are at least moderate.

A type 5 combination consists of two categories that contain minor risk of mortality level diagnoses, plus any two additional secondary diagnoses that are at least a minor risk of mortality level. For a type 5 combination, the minimum patient risk of mortality is moderate. An example of a type 5 combination is atrial fibrillation (category 8) and hypovolemia (category 51) combined with any other two minor secondary diagnoses.

A type 13 combination consists of two secondary diagnosis categories that contain moderate risk of mortality diagnoses, plus any third secondary diagnosis that is at least a moderate risk of mortality diagnosis. For a type 13 combination, the minimum patient risk of mortality subclass is major. An example of a type 13 combination is cirrhosis (category 23) and hypotension (category 50) combined with any other moderate secondary diagnosis.

18. Compute the final risk of mortality subclass

The final patient risk of mortality subclass is computed based on the Phase II base patient risk of mortality subclass and the Phase III modified patient risk of mortality subclasses. If all the Phase III modified risk of mortality are greater than or equal to the Phase II base risk of mortality subclass, then the final risk of mortality subclass is computed as the maximum of the Phase II and III risk of mortality subclasses. If all of the modified Phase III risk of mortality subclasses are less than or equal to the Phase II base risk of mortality subclass, the final risk of mortality subclass is computed as the Phase II base risk of mortality subclass minus one. If the Phase II modified risk of mortality subclasses includes modified risk of mortality subclasses that are both greater and less than the Phase II base risk of mortality subclass, the modified Phase III subclass relating to procedures and combinations of secondary diagnoses will take priority in determining the final risk of mortality subclass. The combination of the APR DRG and the final patient risk of mortality subclass constitute the complete APR DRG description of the risk of mortality of the patient.

Summary of APR DRG risk of mortality subclass assignment logic

The following is a summary of the steps involved in computing the APR DRG risk of mortality subclass of a patient.

Phase I—Determine the risk of mortality level of each secondary diagnosis

  1. Eliminate all secondary diagnoses that are associated with the principal diagnosis of the patient.
  2. Assign each secondary diagnosis its standard risk of mortality.
  3. Modify the standard risk of mortality level of each secondary diagnosis based on the age of the patient.
  4. Modify the standard risk of mortality level of each secondary diagnosis based on the APR DRG and principal diagnosis (applicable only to APR DRG 190 Acute Myocardial Infarct).
  5. Modify the standard risk of mortality level of each secondary diagnosis based on the APR DRG to which the patient is assigned.
  6. Modify the standard risk of mortality level of each secondary diagnosis based on the presence of certain non-OR procedures.

Phase II—Determine the base risk of mortality subclass of the patient

  1. Eliminate all secondary diagnoses that are in the same secondary diagnosis group except the secondary diagnosis with the highest risk of mortality level.
  2. Compute the base patient risk of mortality subclass as the maximum of all the sec�ondary diagnosis risk of mortality levels.
  3. Reduce the base patient risk of mortality subclass if the patient does not have mul�tiple secondary diagnoses at a significant risk of mortality, except for certain secondary diagnoses for which this requirement is removed or modified.

Phase III—Determine the final risk of mortality subclass of the patient

  1. Modify the patient risk of mortality subclass based on the APR DRG and principal diagnosis.
  2. Modify the patient risk of mortality subclass based on the APR DRG and age, or APR DRG and principal diagnosis and age, or APR DRG and birthweight and absence of certain non-OR procedures.
  3. Modify the patient risk of mortality subclass based on a combination of the APR DRG and certain non-OR procedures.
  4. Modify the patient risk of mortality subclass based on the APR DRG and OR procedure.
  5. Modify the patient risk of mortality subclass based on the APR DRG and certain pairs of OR procedures.
  6. Modify the patient risk of mortality subclass based on the APR DRG 583 Neonate With ECMO and the presence/absence of certain OR procedures (this step is appli�cable only to severity of illness, not to risk of mortality).
  7. Modify the patient risk of mortality subclass based upon the APR DRG and princi�pal diagnosis and certain non-OR procedures (this step applicable only to severity of illness, not to risk of mortality).
  8. Establish a minimum risk of mortality subclass for the patient based on the pres�ence of specific combinations of categories of secondary diagnoses.
  9. Compute the final patient risk of mortality subclass based on the Phase II base patient risk of mortality subclass from Step 9 and the modifications of the patient risk of mortality subclass from Steps 10-17.

Conclusion

The APR DRGs form a clinically coherent set of severity of illness and risk of mortality adjusted patient groups. The APR DRGs are designed to describe the complete cross-section of patients seen in acute care hospitals.

Through APR DRGs, hospitals, consumers, payers, and regulators can gain an understanding of the patients being treated, the costs incurred, and, within reasonable limits, the services and outcomes expected. Through APR DRGs, areas for improvement in efficiency and areas with potential quality problems can be identified. The classification of patients into APR DRGs is constantly evolving. As the ICD-9-CM coding scheme changes or as medical technology or practice changes, the APR DRG definitions will continue to be updated to reflect these changes.

Return to Contents
Proceed to Next Section 

Current as of March 2009
Internet Citation: Mortality Measurement: Development of the 3M™ All Patient Refined Diag. March 2009. Agency for Healthcare Research and Quality, Rockville, MD. http://www.ahrq.gov/professionals/quality-patient-safety/quality-resources/tools/mortality/Hughes3.html