population health Flashcards
population health definition
health behaviors/outcomes of a particular population
ex: population could be “Us adults with diabetes”, “children in Camden with asthma”, etc.
this is not the same as public health which is measured to protect the health of everyone
measures of population health should be ____
- timely
-undestandable - have a rigorous methodology
- usable for various populations
- population health measure ≠ clinical quality measure
Examples:
- Geocoding (where individuals live?)
- stratification (sorting individuals by health risk, needs, etc)
- data sources (clinical vs non clinical, insurance claims)
- time interval (how far back?)
- risk adjustment (adjusting resource allocation compensation based on a pts risk for disease complication)
community oriented primary care
an approach where healthcare services are directed at a community’s needs
- define the community
- identify the community’s health issues
- modify the helathcare program in response to the community’s needs
- monitor the impact of program changes
Hot-spotting
identifying patients/communities that heavily use healthcare resources in an attempt to improve their care and lower costs associated with their care
ex. emergency departments
cold - spotting
identifying patients or communities with poor or limited access to healthcare and assessing those limitations for solutions
why is this important?: often patients in cold spots without access to preventative care will progress to hot spot patients which results in more expensive acute care
social deprivation index (SDI)
an overrall measure of social deprivation in an area based on:
1. % of people in area living in poverty
2. % of people in area with <12 years of education
3. % of single parent households in area
4. % of people in area living in rented housing
5. % of people in area in overcrowded housing
6. % of people in area without a car
7. % of adults <65 years old that are unemployed
SDI results in a raw score from 0-100 for a given community where 0 means not disadvantaged and 100 means extremely disadvantaged
impact of risk scoring patients and the concept of risk adjustment
example: medicare and medicaid identify patients with diabetes complications and then adjust their risk score based on social deprivation
- funding/reimbursement adjustments are done in order to allocate resources in a way that addresses social determinants of health
ex, social work allocation, dietician funds - institutions (not individual physicians) are reimbursed for providing these additional services
endemic
baseline prevalence of a disease in a specific community or geographic area
outbreak
sudden spike in disease cases with a community or geographic area
epidemic
a regional disease outbreak (must be within a specific geographical area)
pandemic
internation or worldwide outbreak
ONLY the WHO can declare a pandemic
CDC
center for disease control and prevention
-US federal agency
-can issue travel guidelines
-doesn’t have much authority in and of itself
local health authorities (ex. NJ health department)
-powers vary state by state
-enforce local mask mandates, quarantines, vaccinations, etc.
-perform contact tracing -> data send to CDC
WHO - world health organization
part of the UN
- exclusive authority to declare global public health emergencies such as pandemics
- non-binding international guidelines -> doesn’t have much authority in and of itself
- funded by members of the UN, reliant on UN countries collecting and reporting data -> possibly vulnerable to political pressure
1918 Influenza (H1N1)
-novel H1N1 influenza of unknown origin
Transmissibility:
Ro ≈ 2-3
Morbidity/Mortality:
-mortality ≈ 3-6% of population
-many succumbed to secondary infections
Target Population:
high lethality in 20 - 40 year olds
Therapeutics:
quarantine, masks, social isolation
Misconceptions/Stigma:
known today as “the Spanish Flu” because only Spain was reporting about it as press was slow during World War 1
HIV/AIDS
- retrovirus from animals to humans
- transmitted by blood and body fluids only
Transmissibility:
Ro ≈ 2-5
-asymptomatic transmission
Morbidity/Mortality:
-peak mortality ≈ 16/100,000 in the 1990’s
-slow disease progression
Target population:
disproportionate impact on LGBTQ community and patients with hemophilia, etc
Misconception/Stigma:
a lot of social stigma when mode of transmission was unkown, advocacy from most affected groups played a major role in destigmatization
R naught
R0 – the reproduction number – describes the intensity of an infectious disease outbreak
The formal definition of a disease’s R0 is the number of cases, on average, an infected person will cause during their infectious period.
SARS COV-1 (2002-2003)
-SARS coronavirus
bat origin -> intermediate host -> human
Transmissibility:
Ro ≈ 2
Morbidity/Mortality:
>8k cases, ≈800 deaths
Target Population:
most cases in East/SE Asia
Containment:
intense quarantine intervention by local governments (most cases had severe symptoms so it is easy to identify and impose qurantine)
-this pandemic incited the creation of the WHO
Swine Flu (H1N1) 2009-2010
origin = 2 pig strains
Transmissibility:
Ro < 1.6 (due to available antivirals/vaccines)
Morbidity/Mortality:
60M US cases, <13k deaths
Target Population:
more cases in younger people
Misconceptions/Stigma:
flawed computer models, Tamiflu hoarding
Ebola (2014-2016)
Transmissibility:
Ro ≈1.5-2
Morbidity/Mortality:
~28,000 cases, ~11k deaths in African countries
- lethality rate >50%
Target Population:
most cases in west africa
Therapeutics/Containment:
effective travel screening, rapid drug and vaccine development
COVID 19 (SARS COV-2)
high rate of evolution -> many different strains
-newer strains are less lethal but more transmissible
- interesting fact -> this phenomenon, seemingly paradoxical, makes sense as viruses that kill too fast don’t spread and a virus’s only goal is to spread
- limited immunity from prior infections
Misconception/Stigma:
-Sinophobia, hate crimes against Asians
-Politicization of pandemic
health promotion
actions taken to promote health for ALL people, target is everyone
Primary Prevention
prevent disease in healthy people (example: vaccines)
Target: population that could potentially get disease X
Secondary prevention
early disease detection in sub-clinical patients (examples: colonoscopy)
subclinical: relating to or denoting a disease which is not severe enough to present definite or readily observable symptoms
tertiary prevention
reduce disease severity in symptomatic patients (example: Physical therapy in stroke patients)
US Screening grades
A: Absolutely do this
B: Be a good idea if you did
C: Certain people Could benefit
D: DONT DO IT
I: I dont know bro
epidemiologic curve
From left to right: This shows when the first and last reported illness in the outbreak started.
From bottom to top: This shows how many people got sick on each date during the outbreak.
Shaded area: Some epi curves have a shaded area labeled “Illnesses that began during this time may not yet be reported.” This time frame is known as the “reporting lag,” the amount of time it takes for public health officials to identify and link an illness to an outbreak.
lead-time bias
early screening/diagnosis falsely implies increased survival time
incidence rate (aka incidence density)
(# of new cases) / (people * years of observation)
limitations of incidence:
-may be hard to define new cases
-are new cases really new?
age adjusted mortality
- break your population into age groups (# of people and deaths)
- compare your age-specific mortalities to values from a standardized reference population
- results in a single age-adjusted mortality value for your entire population
quality adjusted life years
a weighted average that accounts for # of years adjusted by the quality of life in that time
Testing, sensitivity, specificity, predictive values and usefulness of tests
Important considerations:
pre-test probability, aka what percentage of people in the population actually do/don’t have the disease, is based on overall prevalence of the disease and/or clinical suspicion
a high pretest probability correlates to a high positive predictive value and a low pretest probability correlates to a high negative predictive value by definition - not that it should matter for the test
Positive predictive value and negative predictive value are heavily influenced by pretest probability, also by definition
Diagnostic accuracy = correct answers/total answers =
(TP+TN) / (TP+TN+FP+FN)
Sensitivity, specificity, and False result rates
-Sensitivity can be thought of essentially “what level can we pick up on a test?”
-Specificity can be thought of essentially “How can we not fuck that up?”
-not being a sensitive enough test, not detecting any fluorescence for example, would mean we get false negative
-not being a specific enough test would mean it could test positive for other things and get a false positive
sensitivity and specificity are inherent to each test design inherently
by intuition, it may seem paradoxical for our sensitivity to rule things out and our specificity to rule things in but tests are made by design such that they have extremely high specificity to prevent false positives and wrong treatment whereas if we have false negatives, we can always rerun the test if the disease etiology is consistent.
likelihood ratios
how much odds change based on a test result
apply intuition we learned while calculating specificity/sensitivity
bayes theorem
describes the connection between conditional probabilities. Bayes theorem dictates that for all partitions of a probability set/space, the sum of all the conditional probabilities equals the probability set/space itself
Fagan’s Diagram
what is this dumb shit??? can we just use numbers?
The Fagan nomogram is a graphical tool for estimating how much the result on a diagnostic test changes the probability that a patient has a disease.
To use this tool, you need to provide your best estimate of the probability of the disease prior to testing. This is usually related to the prevalence of the disease, though this may be modified up or down on the basis of certain risk factors that are present in your patient pool or possibly in this particular patient. You also need to know the likelihood ratio for the diagnostic test.
With this information, draw a line connecting the pre-test probability and the likelihood ratio. Extend this line until it intersects with the post-test probability. The point of intersection is the new estimate of the probability that your patient has this disease.
the computations involved use odds rather than ratios. If you multiply the pre-test odds by the likelihood ratio, you will get the post-test odds. And since multiplication of two numbers is equivalent to adding their logarithms, we use a log scaling for both the odds and the likelihood ratio.
Receiver Operating Characteristic (ROC) curves
Multiple testing - parallel testing
multiple testing - serial testing
