Apache II General Collection Guidelines

When to collect, which value to use

• Select the physiological values that reflect the worst deviation from normal in the first 24 hours in ICU.

• Worst value could be high or low, depending which gives the most points.

Keep in mind the diagnosis and try not to include iatrogenically enhanced symptoms, e.g. BP severely high due to inotropes or decreased LOC due to sedation.

If the length of stay in ICU is less than 24, use those values only.

• NOTE: The Higher the APACHE score, the sicker the patient.

Discussion

• where is this data taken from? Ttenbergen 22:34, 4 June 2008 (CDT)

The following refers to an ongoing discussion, not to a change in collection rules: As discussed in the Critical Care Review Group, the last 4 hours in the ER preceeding admission to the ICU may be admissible data for APACHE values if no data is available for the first 24 hours in ICU. Is this a rule we are using? Ttenbergen 13:12, 17 October 2008 (CDT)

Missing Value Rule

Do not record the value zero and do not leave any blanks.

If a specific value for any APACHE field is not available in the first 24 hours in ICU then: A) Select a value closest to the time prior to admission into ICU. Make a note on your collection sheet where the information was obtained from (ER or Ward etc) and from which date and time period it was obtained from. B) If no information is available, then assume it was not tested because it was normal, and record normal values for all APACHE data. Also note this on your collection in your collection notes. Template:Discussion

Discussion

• there is reference to "the collection sheet" - does that mean we need to track this electronically? If not, why do we need to take this down? Or do we, still?

For patients who "died" in ICU

For patients who "died" in ICU and have no data available in ICU to record for an APACHE select the APACHE data that is closest to the time 'prior to' ICU admission ( values from ER, ward, etc.).

If these are NOT available then:

1. For vital signs (HR, BP mean, RR only), select the lowest limit value on the APACHE table that will give you the highest score.

2. Record normal values for: temperature, Serum CO2, Na, K, Creatinine, HCT and WBC.

Elective Surgery

see Admit Type for APACHE II

Discussion

Used to read:" ELECTIVE SURGERY- YES or NO If surgery is an emergency, or if admission type is NON surgical, please answer NO " Is this inconsistent with the current instructions?

Chronic Health

see... Chronic health field For more info see Chronic Health Diagnosis for APACHE II TOstryzniuk 16:07, 8 December 2008 (CST)

Physiological Variable

Temperature

Select the worst value in the first 24 hours in ICU., i.e. the value that would generate the highest score from the APACHE II scoring table APACHE Scoring table#Chemistry_values.

Discussion

• If the temperature is artificially cooled, should we use these low temperatures? Example when on a cooling blanket or hypothermic from cooling in the OR?--LKolesar 11:58, 19 November 2008 (CST)

mean arterial BP

Look carefully when selecting this value. APACHE II score is based on the worst calculated mean BP (either high or low) which ever gives you the highest points as per scoring table, but you must also take in account the diagnosis. For example if a patient is in septic shock, you would want the lowest BP as opposed to the highest BP that is artificially enhanced by inotropes.

Discussion

• This refers to an arterial BP. Is that different from a regular BP? Ttenbergen 17:14, 25 November 2008 (CST)
• These instructions should be moved to Mean BP and integrated there. Ttenbergen 17:14, 25 November 2008 (CST)

RR (Respiratory Rate)

Select the the value which gives you the highest score regardless of whether a patient is ventilated or not.

HR( Heart Rate)

Select which ever value gives you the highest score. When in atrial fibrillation, do not use this rate, use the ventricular rate.

Discussion

• all are "worst in first 24", right? If so, it is already stated. If stated again, do it for all.

Neurological Assessment

Glasgow Coma Scale

For Head traumas select the worst score in the first 24 hours in ICU. If sedated or paralyzed select the best score in the first 24 hours in ICU prior to sedation. For non neuro & post operative patients who are sedated or paralyzed, please record a normal neuro score, unless there is a neurological problem. If this is the case, you must select the select the worst values prior to sedation or “best guess” based on history prior to sedation. Sedation does not allow us to accurately assess Neuro status, therefore we use alternative information that was documented prior to OR or sedation or we use our “best guess” base on chart notes.

to do before de-stubbing

• what does "best score" mean in this context? Best to make the patient score a high APACHE, or best to let him live?
• in case of sedation or OR, what is the precedence of "alternative" information vs. chart? Are the two not the same?

Blood Gases

Select the worst deviation from normal Arterial Blood Gases (ABG) as outlined on the APACHE II scoring table. Do not use venous gases.

FiO₂ ≤ 49% - Base your selection of the worst ABG on the worst PO2.

FiO₂ ≥ 50% - Base your selection of the worst ABG on the worst AaDO2. AaDO₂ Calculation: (AsDO₂ is the arterial-alveolar O₂ tension difference) AaDO₂ = (Fi02 X 713) - PO₂ - CO₂ e.g. (FiO₂ 0.5 X 713) - PO₂ 85-CO₂ 40 = AaDO₂ 231.5

pH Level

Independ of the ABG selected, chose the worst pH from the blood gases.

No ABG Availabe Rule

If no ABG is available during the first 24 hours in unit use the worst serum CO₂ value.

No ABG and no Serum CO₂ available

Assume normal, record serum CO₂ as 25.0

Laboratory Data

Select the WORST deviation from normal during the first 24 hours in ICU.

• Hematocrit (HCT) is recorded as a percentage, e.g. 0.352 record as 35.2
• Sodium (Na), Potassium (K), Creatinine are recorded as umols
• White Blood Count (WBC) record per 1000 e.g. 10.0

Also see: Missing Value Rule

Legacy Data Information

• Collected for the ICU Program since June 15, 1988 and continues to date.
• Collected for the Medicine Program from October 1, 2003 to December 31, 2006.