Visualizing Multi Organ Failure

Multi Organ Failure (MOF)

Multi organ failure(MOF) or Multi organ dysfunctions syndrome (MODS) is a critical condition where two or more organ system is affected and are not able to function without assistance or intervention . This condition often encountered in patients with sepsis , shock , or trauma . Examples of multi organ failure is congestive heart failure with acute respiratory failure, acute renal failure and liver failure due to septic shock, acute respiratory failure demanding mechanical ventilation when patient in coma. Multi organ failure can be cause due to infection, injury (accident or surgery) or reduced blood flow during acute myocardial infraction or stroke and initial infection trigger an uncontrolled inflammatory response resulting in end -organ damage to vital organs such as heart, lungs, kidney, bowel, brain and liver.

The Problem with Raw ICU Data

The dataset used for this analysis contains the raw data 44 fields and 1552210 rows. Considering the dataset, raw data is high in volume and complex data set .Clinician track values such as Lactate, Platelets , Creatinine , Billiuirubin, PaO2/FiO2 ratio , Blood pressure ,Heart rate , respiratory rate . During analysis without any structure identifying change over from organ stress to organ failure.

Raw Data to Clinical Insights Using MOD SCORE

The Multi Organ Dysfunction (MOD) Score converts multiple organ biomarker into a single clinical indicator .

Kidney ,respiratory,liver,Mod score parameters are merged together and calculate mod score . Mod score allows clinicians to detect organ failure severity and also it helps to track deterioration over time,predict mortality rate.

Visualizing Organ Failure

Organ failure is visualized using tools like Tableau is where raw data is converted into powerful analyzing tool. Tableau dashboards can be used to overview realtime patient condition.

 High Risk Patients By Gender

Calculated Field Used

1. High Risk Patients

SUM(IF [MOD_Score] >= 6 THEN 1 ELSE 0 END)/ COUNT([Patient ID])

2.GENDER CATEGORY

IF [Gender] = 0 THEN 'Female'

ELSEIF [Gender] = 1 THEN 'Male'

ELSE 'Unknown'

END

High Risk Patients By Gender help us to detect the patient who need immediate intervene , were being in a critical condition by gender wise.

Trend Line of Multi Organ dysfunction

Calculated Field Used

1. Organ Dysfunction

IF { FIXED [Patient ID] : MAX(

IF [Troponin I] > 0.04 THEN 1 ELSE 0 END +

IF [Gender] = 0 THEN

IF [Creatinine] < 0.6 OR [Creatinine] > 1.1 THEN 1 ELSE 0 END

ELSE

IF [Creatinine] < 0.7 OR [Creatinine] > 1.3 THEN 1 ELSE 0 END

END

)} >= 1 THEN 'HEART DYSFUNCTION'

ELSEIF { FIXED [Patient ID] : MAX(

IF [AST] <10 OR AST > 40 THEN 1 ELSE 0 END +

IF [Alkalinephos] <44 OR [Alkalinephos] > 147 THEN 1 ELSE 0 END +

IF [Biliirubin total] >= 1.2 THEN 1 ELSE 0 END +

IF [Billirubin direct] > 0.3 THEN 1 ELSE 0 END

)} >= 1 THEN 'LIVER DYSFUNCTION'

ELSEIF { FIXED [Patient ID] : MAX(

IF [Potassium]<3.5 OR [Potassium]>5.2 THEN 1 ELSE 0 END +

IF [Chloride]<96 OR [Chloride] > 106 THEN 1 ELSE 0 END +

IF [Calcium] < 8.6 OR [Calcium]>10.3 THEN 1 ELSE 0 END +

IF [Gender] = 0 THEN

IF [BUN] < 6 OR [BUN]>21 THEN 1 ELSE 0 END

ELSE

IF [BUN] < 8 OR [BUN]>24 THEN 1 ELSE 0 END

END +

IF [Gender] = 0 THEN

IF [Creatinine] < 0.6 OR [Creatinine] > 1.1 THEN 1 ELSE 0 END

ELSE

IF [Creatinine] < 0.7 OR [Creatinine] > 1.3 THEN 1 ELSE 0 END

END +

IF [Glucose] > 140 THEN 1 ELSE 0 END

)} >= 1 THEN 'KIDNEY DYSFUNCTION'

ELSEIF { FIXED [Patient ID] : MAX(

IF [P H] < 7.35 OR [P H] > 7.45 THEN 1 ELSE 0 END +

IF [Base Excess] < -2 OR [Base Excess] > 2 THEN 1 ELSE 0 END +

IF [FiO2] < .21 THEN 1 ELSE 0 END +

IF [SaO2] <95 OR [SaO2]>99 THEN 1 ELSE 0 END +

IF [PaCO2] < 38 OR [PaCO2] > 42 THEN 1 ELSE 0 END +

IF [Resp] < 12 OR [Resp] > 20 THEN 1 ELSE 0 END

)} >= 1 THEN 'LUNG DYSFUNCTION' END

To calculate Organ dysfunction ,Firstly the value of troponin and creatinine value is not normal then it is considered as Heart dysfunction. Next for liver dysfuntion the biomarker AST,Alkalinephos,Billirubin direct and Billirubin total are considered together and if they are abnormal .Third organ is kidney dysfunction has a group of markers such as potassium, chloride, calcium, bun of all values should be abnormal for male and for female the other markers are considered such as bun , creatinine , glucose are abnormal .Last Lung dysfunction may have abnormal values of ph, base excess, fio2, sao2, paco2 and respiratory rate.

2. MOD SCORE

[MOD _Billirubin Score]+[MOD _Creatinine_Score]+[MOD _Creatinine_Score]+[MOD_PaO2/FiO2]+[MOD_Platelets_Score]

3.AVG. MOD SCORE

Avg(MOD SCORE)

MOD Score is a score which is calculated by using the value of billirubin ,creatinine , platelets and pao2/fio2 each have the value one if the values are abnormal. In next step every count is added together , then count is more than 2 it will be a mod score. Using the calculation organ dysfunction is detected in the chart. Its clear visualization shows timely detection of the organ failure encounters.

Multi Organ Failure by MOD SCORE

Calculated Field Used

1. Organ dysfunction

2.High risk patients

3 . Gender Category

4. MOD Score

5. Age Category

IF [Age]<=15 THEN "0-15"

ELSEIF [Age]>15 AND [Age]<=30 THEN "15-30"

ELSEIF [Age]>30 AND [Age]<=50 THEN "31-50"

ELSEIF [Age]>50 AND [Age]<=75 THEN "51-75"

ELSEIF [Age]>75 AND [Age]<=100 THEN "76-100"

END

Multi Organ Dysfunction in a tree map helps clinician to detect patients segregate into smaller group by age and gender. Age category in this chart is grouped from the age of (0-15 ) in a one group, age (15-30) as second group, age(30-50) as third group , age (50-75) as fourth group and age (75-100) as a another group are using if else condition statement. Segregation used here are helpful to clinicians decide timely and rapidly treating patients .

Visualization Improves Outcomes

Well planned visual dashboard help clinicians to detect early deterioration. It helps to prioritize high risk patients and also in deciding treatments. Because of timely intervention mortality rate is reduced . Instead of reacting to failure, teams begin preventing it .

Conclusion

In analysis of sepsis patients , visualizing MOD Score trends witnessed that patients scores elevated rapidly within first 24 hours had significantly higher mortality. Multi organ failure is a race against time .when raw data is transformed into visual intelligence,clinicians gain the power to intervene earlier, and faster.