L57. EBM - Prognosis, Risks & Odds

• Objectives:
  • Define prognosis, morbidity, and mortality
  • Describe ways to express prognosis
  • Demonstrate proficiency in calculating risk estimates (RRR, ARR, & NNT)
  • Understand the difference between Risk Ratio and Odds Ratio and their calculations

Morbidity

• CDC defines morbidity as any departure, subjective or objective, from a state of physiological or psychological well-being, including disease, injury, and disability.
• Morbidity refers to the the number or persons who are ill, illness periods, or illness durations.
• Data comes from disease registries and morbidity surveys like NHANES and the Framingham Heart Study.

Mortality

• CDC defines mortality as death, and mortality rate as the frequency of death in a defined population during a specified interval.
• Morbidity and mortality measures are often mathematically similar, differing by what is being measured (illness or death).
• Data sources: Death certificates, hospital records, and autopsy data.

Frequently Used Measures of Morbidity

• Incidence proportion (attack rate/risk): new cases of disease during a specified time interval/population at the start of the time interval.
• Secondary attack rate: number of new cases among contacts/total number of contacts.
• Incidence rate (person-time rate): number of new cases of disease during a specified time interval/summed person-years of observation or average population during time interval.
• Point prevalence: number of current cases (new and preexisting) at a specified point in time/population at the same specified point in time.
• Period prevalence: number of current cases (new and preexisting) over a specified period of time/average or mid-interval population.

Frequently Used Measures of Mortality

• Crude death rate: total deaths during a given time interval/mid-interval population.
• Cause-specific death rate: deaths assigned to a specific cause during a given time interval/mid-interval population.
• Proportionate mortality: deaths due to a specific cause during a given time interval/total deaths from all causes during the same time interval.
• Death-to-case ratio: deaths assigned to a specific cause during a given time interval/new cases of same disease reported during the same time interval.
• Neonatal mortality rate: deaths among children < 28 days during a given time interval/live births during the same time interval.
• Postneonatal mortality rate: deaths among children 28-364 days during a given time interval/live births during the same time interval.
• Infant mortality rate: deaths among children < 1 year during a given time interval/live births during the same time interval
• Maternal mortality rate: deaths assigned to pregnancy-related causes during a given time interval/live births during the same time interval.

Prognosis

• Prognosis predicts a disease's course, duration, and outcome based on its pathogenesis and risk factors.
• Prognosis is established after diagnosis and before the treatment plan.
• Case-Fatality Rate (CFR): (# of people who die from the disease)/(# of people with the disease) = %
  • It's often used for acute diseases of short duration.
  • In chronic disease, death may occur long after diagnosis making the possibility of death due to another cause more likely.
• Example: The SARS 2003 Epidemic lasted from February to June, with 8098 cases reported and 774 deaths
  • CFR = 774/8098 = 9.6%

Kaplan-Meier Curve

• Identifies the exact time an event (death or lost to follow up) occurs.
• Graphic representation of an event's probability over time, demonstrating survival probability = 1 – (event probability).

Survival

• Median Survival Time: The duration time when half the study population is still surviving.
  • This has advantages over mean survival:
  • It's less affected by extremes (outliers).
  • It can be calculated before the end.
• Survival Rate: represents % of people alive at a certain time after diagnosis
  • Ex: 5 Year, Disease-free, Progression-free
  • Influenced by type of cancer, stage, age and time for cancer cases.

Relative Risk (RR)

• RR = (Event Rate in Intervention Group) / (Event Rate in Control Group)
• Relative Risk Reduction (RRR) = 1 - relative risk, or (Absolute Risk Reduction) / (Event rate in control group)
• Absolute Risk Reduction (ARR) = (Event rate in control group) - (Event rate in intervention group)
• Number Needed to Treat (NNT) = 1 / (absolute risk reduction)
• Relative risk reduction is often more impressive than absolute risk. The lower the event rate in the control group, the larger the difference between relative risk reduction and absolute risk reduction.

Benefit and Harm Table

• Number Needed to Treat: number of patients who would need to receive treatment for one to benefit

  • calculated as 100 divided by the absolute risk reduction expressed as a percentage (or 1 divided by the absolute risk reduction expressed as a proportion)

• Number Needed to Harm: number of patients that would have to receive the treatment for one of them to experience an adverse effect

  • calculated by 100 divided by the absolute risk increase expressed as a percentage (or 1 divided by the absolute risk increase expressed as a proportion)

Examples of Calculations

• A new drug decreases the Heart Attack rate by 50%

  • 2000 women, ages 30-35 who are non-smokers and w/ no other medical problems are randomized to receive this new drug, then followed for 6 months.
  • Rate of MI:
    • Control Group 2/1000
    • Treatment Group 1/1000
  • RR = (1/1000)/(2/1000) = 0.5
  • RRR = 1 - 0.5 = 0.5 = 50% Relative Risk Reduction
  • ARR = 0.002 - 0.001 = 0.001 = 0.1% Absolute Risk Reduction
  • NNT = 1/ARR = 1/0.001 = 1000
  • Need to treat 1000 patients to prevent one MI

• A group of 2000 men, ages 60-65 with hypertension and hyperlipidemia are randomized to receive this new drug, then followed for 12 months.

  • Rate of MI:
    • Control Group 60/1000
    • Treatment Group 30/1000
  • RR = (30/1000)/(60/1000) = 0.5
  • RRR = 1 - 0.5 = 0.5 = 50% Relative Risk Reduction
  • ARR = 0.06 - 0.03 = 0.03 = 3% Absolute Risk Reduction
  • NNT = 1/ARR = 1/0.03 = 33
  • Need to treat 33 patients to prevent one MI.

• The drug has a potential side effect of causing acute renal failure. Among the men randomized above, over 12 months the rate of acute renal failure was

  • ARF:
    • Control Group 10/1000
    • Treatment Group 20/1000
  • RR = 2
  • Relative Risk Increase 100%
  • ARI = 0.01 = 1%
  • NNH = 1/0.01 = 100
  • If you treat 100 patients, 1 will develop acute renal faliure.

Attributable Risk (ATR)

• ATR is the risk difference between exposed and unexposed groups.
• For example, the risk of breast cancer in the general population is about 13%. The risk is about 60% if a person has the BRCA1 gene mutation.
• ATR = 60% - 13% = 47% = 0.47

Risk Ratio

• Compares probabilities in two groups, also known as the risk ratio; ratio of two conditional probabilities.
• An RR = 1 implies the event is equally probable in both groups.
• An RR > 1 implies that the event is more likely in the first group, and an RR < 1 implies that the event is less likely in the first group.
• Often used for RCT and cohort studies.
  
• RR = a/(a+b) / c/(c+d)

Odds Ratio

• Compares the probability of an event between 2 groups.
• An odds ratio of 1 implies that the event is equally likely in both groups.
• An odds ratio greater than one implies that the event is more likely in the first group.
• An odds ratio less than one implies that the event is less likely in the first group.
• Often used for case-control studies, because the total denominator for the entire population is unknown.
• OR = a/b / c/d
• Which can be simplified to OR = ad/bc
• If the odds are the same in each row, then the odds ratio is 1.

Additional Points for Risk Ratio and Odds Ratio

• These measures do not indicate the absolute probability that someone exposed to the risk factor will develop the disease.
• If the frequency of a disease is 2/million in the exposed group, but 1/million in the unexposed, RR is doubled by exposure if the groups are of equal size. However, the disease is still rare in both groups.
• For uncommon events (e.g., frequency < 10% of the population), the values of RR and OR are nearly the same.
• For common events, the OR may increase disproportionately while that of RR is constrained. OR is always further away from 1 than RR. If exposure increases risk, then OR > RR > 1.
• RR or OR should be presented with the corresponding 95% confidence interval (CI). If both the lower and upper confidence limits are > 1, than the exposure increases the risk. If both limits are < 1, then the exposure is protective. However, if the confidence interval includes 1, then you cannot conclude either way.