Lecture 3 and_4_Measures_of_Disease_Occurence.pdf

Lectures 3 & 4 Measures of Disease Occurrence Prof. Shafika Assaad 1 “We owe all the great advances in knowledge to those who endeavor to find out how much there is of anything”. ~James Maxwell, Physicist, 1831‐1879 ...

Lectures 3 & 4 Measures of Disease Occurrence Prof. Shafika Assaad 1 “We owe all the great advances in knowledge to those who endeavor to find out how much there is of anything”. ~James Maxwell, Physicist, 1831‐1879  Measures of Disease Occurrence In order to study the distribution and determinants of a certain outcome, it is first necessary to measure its frequency. Frequency is the rate at which some thing occurs or is repeated over a particular period of time or in a given sample Several mathematical measures are used to convey information about the occurrence of disease. These include: – Counts – Proportions – Ratios Used interchangeably sometimes, be careful! – Rates 3  Counts A count refers to the number of cases of the disease or health‐related event under investigation. Simplest and most frequently used measure. Example: – Number of cases of influenza in Beirut in February 2012. – Number of injuries related to car accident in Lebanon. – College dorm students who had hepatitis in a certain year 4  Counts Limitation When we take the population in account, Number of Population disease occurrence is: Cases of Total – City A= number of cases/ total Disease population= 25/50= 0.5 = 50% City A 25 50 – City B= 50/200= 0.25 = 25% Counts erroneously indicate that the disease is more common in City B than A City B 50 200 For epidemiologic purposes, counts are not very informative without knowing the population size – Cannot make comparisons between populations based on counts Useful for health resources allocation (personnel and services needed to treat the sick people) within each population5  Proportion A proportion indicates the Proportion= A/ A+B fraction of a population = Number of people with that is affected with a disease (A)/ total number certain condition of population Persons included in the Takes a numeric value numerator are always between 0 and 1 part of the denominator Can be expressed as a percentage 6  Proportion Example A B Number of men Number of men with CVD without CVD 517 974 Proportion of men with CVD = ?\ 7  Proportion Example A B A + B (Total) Number of men Number of men Total number of with CVD without CVD men in study population 517 974 1491 Proportion of men with CVD = A/ A+B = 517/ 1491 = 0.347 = 34.7% 8  Ratio In a ratio, the numerator and the denominator are independent (they have no specific relationship) i.e. persons included in the numerator are not part of the denominator Ratio= A/B 9  Ratio (Refer to Previous Example) A B A + B (Total) Number of men Number of men Total number of with CVD without CVD men in study population 517 974 1491 ▪ Ratio of men without CVD to men with CVD= = B/ A= 974/ 517 = 1.88 :1 (~ 2 to 1) ▪ For every 2 men who don’t have CVD, one man has the disease 10  Ratio Example Sex Ratio for AIDS mortality: = number of male deaths (X)/number of female deaths (Y) X= 450,451 and Y= 89,895 Sex ratio would be approximately 5:1 The number of men who die from AIDS is approximately 5 times higher than the number of women who die from AIDS 11  Rate A rate is a ratio that incorporates the element of TIME. It consists of a numerator and denominator in which time forms part of the denominator Includes the following components: – Disease frequency (numerator) – Population size – Time period during which an event occurs – For example, in the United States in 2002, a total of 15,075 new cases of tuberculosis were reported. (3) During the same year, 802 deaths were attributed to tuberculosis. The tuberculosis death-to-case ratio for 2002 can be calculated as 802 ⁄ 15,075. E.g. 11 cases of tuberculosis per 100,000 12persons per year Measures of Morbidity  Measures of Morbidity Measures of morbidity- Having a disease or symtoms of a disease Incidence Prevalence 14  ‫ حدوث‬Incidence Refers to the development of NEW cases of a disease during a defined period of time in previously disease‐free individuals who are at risk of developing the disease (i.e. capable of developing the disease) Incidence quantifies the development of disease, i.e., the transition from a non‐diseased to diseased state 15  Incidence Provides an estimate of the probability (or risk) that a person will develop a disease during a specified period of time Central to study causal factors of diseases 16  Incidence Rate New cases occurring during a given time period Incidence Rate= X10 n Population at risk during the same time period – Any individual counted in the denominator must have the potential to become a case – Example: If we are calculating the incidence of ovarian cancer: We CANNOT include men in the denominator Men are NOT at risk for developing ovarian cancer Women who have had oophorectomy (surgical removal of both ovaries) would not be at risk of developing ovarian cancer and should be excluded from denominator 17  Types of Incidence Measures Cumulative incidence – Presented on previous slide Incidence density – Will be discussed later in the course 18  Prevalence - Entishar Refers to EXISTING cases of disease or health condition in a population (generally irrespective of the duration of disease) Quantifies the burden or “magnitude” of disease Who has the disease and who doesn’t regardless of when the disease developed (it could have developed yesterday, last month, last year, or 10 years ago) “Snapshot” of disease in the population Prevalence Rate= Number of cases of a disease present in the population at a specified time X10n Number of persons in the population at that specified time 20  Prevalence 2 types of prevalence measures Point prevalence Period prevalence 21  Point Prevalence Number of existing Example: cases /total population On September 1, 2012, at a set point in time Community A had: (e.g. September – Population= 1600 1,2012) – Existing cases of influenza= 29 Point prevalence of influenza= 29/ 1600 = 1.8% 22  Point Prevalence Point Prevalence=Number of persons ill at a time point Total number in the group In Columbia, students were asked “Do you smoke cigarettes now? The total number in the group was 41,837.The total number who responded Yes was 6234. Therefore , the prevalence of current smokers in Columbia on January 16, 2020 was: 6234 41837 This result could be expressed as a percentage 14.9% or as a frequency per 1000 ( 149.0)  Period Prevalence Number of existing cases Example: (refer to previous /total population during a slide) specified time period (e.g. –Between Sep 1, 2012 and Sep 1, 2012 to Sep 30, Sep 30, 2012, Community A has: 2012) Population= 1600 (let’s assume population remains fixed) Existing cases of influenza on Number of cases= existing Sep 1, 2011= 29 cases on Sep 1 (irrespective New (Incident) cases of of disease duration), AND influenza between Sep 1 and Sep 30= 7 those newly diagnosed until Sep 30, 2012 – Period prevalence= (29+7)/ 1600= 0.0225= 2.25% 22  Period Prevalence Period Prevalence= number of persons ill during a time period Average population Women were asked: Have you ever been diagnosed by a physician as having any form of cancer, other than skin cancer? Note that the question did not ask about current disease but rather about lifetime history. Thus it refers to period prevalence , the period being the entire life span. To calculate the period prevalence , one needs to know the average population( still 41,837) and the number who responded Yes ( 2293). Therefore the period prevalence of cancer in the study population was : 2293 x 100 41837 5.5% Incidence and Prevalence Illustration Case A Case B Case C Case D Case E January 2012 June 2012 December 2012 Represents disease duration (from time of diagnosis till time of recovery (or death) 23  Incidence and Prevalence Illustration What is the period prevalence for Jan 2012‐ Dec 2012? What is the incidence for 2012? What is the point prevalence for: – Jan 2012? – June 2012? – Dec 2012? **Note: For purposes of this illustration, ignore the denominator, let’s just focus on quantifying the numerator 27  Answers What is the period prevalence for Jan 2012‐Dec 2012? – Answer: Numerator will include 5 cases (A, B, C, D, and E) What is the incidence for 2012? – Answer: 2 cases (B and C) What is the point prevalence for: – Jan 2012? Answer: 3 cases (A, D, and E) – June 2012? Answer: 4 cases (B, C, D, and E) – Dec 2012? 28 Answer: 2 cases (B and E)  Identify Type of Measure Question Type of Measure Do you currently have hypertension? Have you had hypertension during the past (n) years? Have you received a hypertension diagnosis in the last year? 29  Identify Type of Measure Question Type of Measure Do you currently have Point prevalence hypertension? Have you had hypertension during Period prevalence the past year? Have you received a hypertension Incidence diagnosis in the last year? 30  Incidence vs. Prevalence Incidence Prevalence Focuses on New events Focuses on EXISTING events Measures the risk of Measure of the magnitude developing disease; the rate or “burden” of disease in a at which disease develops per population year (or unit of time) in a Think about it as the population odometer of a car (miles Needed to investigate accumulated within a certain disease etiology period) Very helpful in planning ---Think about it as the health services speedometer of a car (60 28 miles per hour)  How to determine Prevalence and Incidence? Prevalence Incidence Existing NEW “Prevalent” “Incident” Cases! Cases! Have the Developed the disease disease Defined Do not have Population the disease Do not have Did not the disease develop the disease Follow up in time (x yrs) Start with a defined population at a Take those who did not have the specified point in time “baseline” disease at “baseline” Screen the population to determine who Follow them up for a certain period of has the disease and who does not have it time (e.g. 1 year) and rescreen them to determine who DEVELOPED the disease 29  Hypothetical Example: Uterine Cancer Prevalence Incidence 150 developed 50 had uterine Population of cancer Women without uterine cancer postmenopausal uterine cancer women, age 50‐79 89,950 89,800 remained years 89,950 did not have the disease free N= 90,000 disease Follow up for 8 years Study began in January 2000 Women who did not have uterine cancer were Women were asked: “have you had uterine followed up for a period of 8 years to ascertain the cancer during the past 5 years”? development of uterine cancer Question 1: What is the prevalence of uterine Question 1: what is the incidence rate of uterine cancer? cancer over 8 years? Question 2: What type of prevalence measure Question 2: Assume that 9950 women out of 89950 is this? were found to have had hysterectomy. Does this affect the incidence rate? 30  Example: Uterine Cancer (Answers) Prevalence Incidence 50 have uterine 150 developed Population of cancer postmenopausal Women without uterine cancer women, age 50‐79 uterine cancer years 89,950 do not 89,950 89,800 remained N= 90,000 have the disease free disease Follow up for 8 years Question 1: Question 1: Incidence of uterine cancer= 150/89,950= Prevalence of uterine cancer= 0.00166= 166 per 100,000 (per 8 yrs) 50/90,000= 0.000556= 55.6 per 100,000 Question 2: Question 2: ‐ Women with hysterectomy should be Period prevalence (Jan 1995‐Jan 2000) excluded from denominator. ‐ Incidence rate= {150/(89,950‐9950)}= 34 0.00189= 189 per 100,000  Interrelationship between Incidence and Prevalence Incidence and prevalence are interrelated – When the incidence of a disease increases, its prevalence in the population also increases Consider 2 diseases A and B: – They have the same incidence rate – A has long duration and B has short duration – Which disease has a higher prevalence? Answer= disease A – When the duration of a disease increases, its prevalence increases 35 Interrelationship between Incidence and Prevalence (cont’d) Prevalence=Incidence x Duration of Disease Assuming: A steady‐state situation: Incidence rate is constant over time Duration of disease is constant over time 36 Factors that Influence Prevalence Factors that increase prevalence: ‐ Increase in disease incidence ‐ Longer duration of disease ‐ Treatments that prolong the patient’s life but do not cure the disease ‐In‐migration of cases ‐In‐migration of people at risk of developing the disease ‐ Better diagnosis Factors that decrease prevalence: ‐ Decrease in disease incidence ‐ Shorter duration of disease ‐ High death rate from disease ‐ Treatments that cure disease ‐ Out‐migration of cases 37  Incidence and Prevalence Shete jdid http://painconsortium.nih.gov/symptomresearch/chapter_19/sec4/cihs4pg1.htm 36 Measures of Mortality Common Measures of Mortality Annual mortality rate Case Case fatality fatality rate rate Proportionate mortality rate Measures of Infant mortality rate Mortality Maternal mortality ratio Life expectancy Years of Life lost, etc.. 38  Annual Mortality Rate Total number of deaths from all causes in 1 year Number of persons in the population at mid year It is standard practice to take the size of the population at mid-year as the denominator because population size may vary during the year (due to migration, births and deaths) and the mid- year population serves as an estimate of the average population exposed to the risk of dying over the Represents the “risk” of dying from any cause in a given year Denominator: includes population from which deaths occurred Because the population changes over time, an approximation is used: – Generally, we use the number of persons in the population at midyear Usually multiplied by 10n to get rid of decimals (e.g. multiplied by 103 and expressed per 1,000 population) 39  Annual Mortality Rate: Example Death rate in the US in the year 2005: Number of deaths in the US during 2005= 2,448,017 Population of the US on July 1,2005 (midyear)= 296,410,404 Annual death rate in 2005: = 2,448,017/ 296,410,404 = 0.008259 = 8.259 per 1,000 (multiply by 103) = 825.9 per 100,000 (multiply by 105) 40  Mortality Rate Mortality rate does not have to be annual Can be calculated over 5 years, etc.. Time period is arbitrary BUT must be precisely specified 43  Case Fatality Rate Number of deaths due to certain disease during a time period Total number of disease cases during same period Number of deaths due to a disease that occur among people who have that disease Measure of “lethality” and “severity” of a disease Links mortality to morbidity Usually multiplied by 100 and expressed as % 44  Case Fatality Rate: Example Hantavirus infection: 45 cases of hantavirus infections occurred in a US state in the year 1993 22 cases out of 45 were fatal Case Fatality Rate (CFR): = (22/45)*100= = 48.9% Interpretation: approximately 49% of people who get a hantavirus infection die from it (deadly infection) 45  Proportionate Mortality Rate Mortality due to a specific cause during a time period Mortality due to all causes during same period What proportion of all deaths is caused by a specific disease? Usually multiplied by 100 and expressed as % 46  Proportionate Mortality Ratio: Example Leading causes of death in the US, 2005 Rank Cause of Death Number of Deaths Proportionate Mortality Ratio ‐‐‐ All Causes 2,448,017 100% 1 Heart disease 652,091 26.6% 2 Cancer 559,312 22.8% 3 Cerebrovascular disease 143,579 5.9% 4 Chronic lower respiratory 130,933 5.3% disease 5 Accidents 117,809 4.8% 6 Diabetes mellitus 75,119 3.1% 7 Alzheimer’s disease 71,599 2.9% 8 Pneumonia and influenza 63,001 2.6% 47  Proportional Mortality Rate for Heart Disease in 2005 Deaths due to heart disease in 2005 *100 Total number of deaths in 2005 = (652,091/2,448,017)*100 = 26.6% Interpretation: of all deaths that occurred in the US in 2005, 26.6% were caused by heart disease Note: If the PMR from heart disease increases (over time for e.g.), this does not mean that the risk of death from heart 46 disease is increasing  Comparison of Proportionate Mortality Rate (PMR) and Mortality Rate 1990 2010 Mortality rate from all 30/1,000 15/1,000 causes in Community X PMR from heart disease 10% 20% in Community X Can we conclude based on the PMR that the risk of dying from heart disease in Community X doubled between 1990 and 2010? NO! Let’s do the math and calculate heart disease mortality rates in 1990 and 2010 49  Comparison of PMR and Mortality Rate 1990 2010 Mortality rate from all 30/1,000 15/1,000 causes in Community X PMR from heart disease 10% 20% in Community X Mortality rate from heart 3/1,000 (10% 3/1,000 (20% disease of 30/1,000) of 15/1,000) Mortality rates from heart disease in Community X remained stable between 1990 and 2010 i.e. the risk of dying from heart disease in both time periods is the same 50  Comparison of PMR and Mortality Rate Conclusion: PMR is not a measure of risk of dying from a certain cause Do not confuse it with the Mortality Rate for that cause If the PMR for heart disease increases over time, it does not necessarily mean that the risk of death from heart disease is also increasing – It could mean that mortality from other causes (e.g. cancer) is decreasing and cancer is contributing less to overall mortality 51  Infant Mortality Rate Number of infant deaths among infants aged 0‐365 days in a given year Number of live births during same year Usually expressed per 1,000 live births Powerful indicator of health status of country Infant Mortality Rate, 2010 Lebanon 19 infant deaths per 1,000 live births Afghanistan 103 infant deaths per 1,000 live births Sweden 2 infant deaths per 1,000 live births USA 7 infant deaths per 1,000 live births 50  Years of Potential Life Lost (YPLL) Measure of premature mortality. The concept of years of potential life lost (YPLL) involves estimating the average time a person would have lived had he or she not died prematurely Age of death for every deceased person is subtracted from a predetermined “standard” age at death (e.g. 65 years in the US) – YPLL for an infant dying at 1 year= (65‐1)= 64 years – YPLL for a man dying at 50= 15 year – Younger age of death more years of potential life lost Total YPLL is calculated by adding all individuals’ “years of potential life lost” for each cause of death Years of potential life lost (YPLL) before age 65, all races, both sexes, all deaths, United States, 2004 54  Different Forms of Rates Rates used to quantify morbidity and mortality can be of different forms: 1.Crude: – Calculated for a population as a whole – Does not take in account differences in population structure such as age and sex – Unadjusted Crude death rate= number of deaths in a given year x100000 population- mid year in 2003 55 In USA number of deaths= 2448288- population as of July=290810789 Crude death rate= 2448288/290810789=841.9 per 100000  Different Forms of Rates (cont’d) 2. Specific or Stratified: – Calculated for specific subgroups within a population or for specific causes of death/ diseases: Cause‐specific rate Age‐specific rate Sex‐specific rate Race‐specific rate 3. Adjusted or Standardized: – Statistically calculated rate that is adjusted according to a standard population 56  Cause‐specific Rate Number of deaths (or number of cases) due to a certain disease Population size at midpoint of time period Usually multiplied by 100,000 Example: Go back to the table of causes of death in the US in 2005 Let’s calculate the cause specific‐mortality rate for accidents 55 Leading causes of death in the US, 2005 Rank Cause of Death Number of Deaths ‐‐‐ All Causes 2,448,017 1 Heart disease 652,091 2 Cancer 559,312 3 Cerebrovascular disease 143,579 4 Chronic lower respiratory disease 130,933 5 Accidents 117,809 6 Diabetes mellitus 75,119 7 Alzheimer’s disease 71,599 8 Pneumonia and influenza 63,001 56 Accidents‐specific Mortality Rate Number of deaths due to accidents in 2005 Population size in the year 2005 Number of deaths due to accidents= 117,809 Population size on July 1, 2005 (selected at midpoint of the year)= 296,410,404 Accidents‐specific mortality rate = 117,809/296,410,404 = 0.000397 Or 39.7 per 100,000 57  Accidents‐specific Mortality Rate Interpretation: 39.7 per 100,000 people in the US in the year 2005 died because of accidents 60  Age‐specific Mortality Rate Number of deaths within a certain age group Number of people in the same age group (during same period) Usually multiplied by 100,000 Example: In the US in 2005, there were 1,717 deaths due to cancer in the age group 15‐24 years. Number of people in the 15‐24 years age group= 42,076,849. What’s the age‐specific cancer mortality rate? 61  Age‐specific Mortality Rate (cont’d) The age‐specific cancer mortality rate in the 15‐24 years age group is: (1,717/42,076,849)*105 = 4.1 per 100,000 62 More Examples: Cause‐specific Mortality Rate Table. Distribution of Myocardial Infarction (MI) in 2 counties in South Carolina, 1978 Population Subgroup Population Size Number of MI Deaths due to MI cases White men 17,902 130 16 White women 20,142 35 7 Black men 8,832 17 3 Black women 11,253 11 1 Total 58,129 193 27 Cause‐specific mortality rate due to MI= 63  More Examples: Cause‐specific Mortality Rate Table. Distribution of Myocardial Infarction (MI) in 2 counties in South Carolina, 1978 Population Subgroup Population Size Number of MI Deaths due to MI cases White men 17,902 130 16 White women 20,142 35 7 Black men 8,832 17 3 Black women 11,253 11 1 Total 58,129 193 27 Cause‐specific mortality rate due to MI= = (MI deaths/ population total)*105 = (27/ 58,129)*105 = 46 deaths per 100,000 residents 64  More Examples: Case Fatality Rate Table. Distribution of Myocardial Infarction (MI) in 2 counties in South Carolina, 1978 Population Subgroup Population Size Number of MI Deaths due to MI cases White men 17,902 130 16 White women 20,142 35 7 Black men 8,832 17 3 Black women 11,253 11 1 Total 58,129 193 27 MI Case‐fatality rate: 65  More Examples: Case Fatality Rate Table. Distribution of Myocardial Infarction (MI) in 2 counties in South Carolina, 1978 Population Subgroup Population Size Number of MI Deaths due to MI cases White men 17,902 130 16 White women 20,142 35 7 Black men 8,832 17 3 Black women 11,253 11 1 Total 58,129 193 27 MI Case‐fatality rate: = (deaths due to MI/ number of MI cases)*100 = (27/193)*100 = 14% i.e. 14% of individuals who suffered from MI, died as result of 66 it  More Examples: Sex‐specific Mortality Rate Table. Distribution of Myocardial Infarction (MI) in 2 counties in South Carolina, 1978 Population Subgroup Population Size Number of MI Deaths due to MI cases White men 17,902 130 16 White women 20,142 35 7 Black men 8,832 17 3 Black women 11,253 11 1 Total 58,129 193 27 Sex‐specific MI mortality rate: 67  More Examples: Sex‐specific Mortality Rate Table. Distribution of Myocardial Infarction (MI) in 2 counties in South Carolina, 1978 Population Subgroup Population Size Number of MI Deaths due to MI cases White men 17,902 130 16 White women 20,142 35 7 Black men 8,832 17 3 Black women 11,253 11 1 Total 58,129 193 27 Sex‐specific MI mortality rate: ‐ MI mortality rate in men: = (MI deaths in men/ number of men)*105 = (16+3)/ (17,902+8,832) = 71 deaths per 100,000 ‐ Also calculate rate in women 68  More Examples: Race‐specific Mortality Rate Table. Distribution of Myocardial Infarction (MI) in 2 counties in South Carolina, 1978 Population Subgroup Population Size Number of MI Deaths due to MI cases White men 17,902 130 16 White women 20,142 35 7 Black men 8,832 17 3 Black women 11,253 11 1 Total 58,129 193 27 Race‐specific MI mortality rate: 69  More Examples: Race‐specific Mortality Rate Table. Distribution of Myocardial Infarction (MI) in 2 counties in South Carolina, 1978 Population Subgroup Population Size Number of MI Deaths due to MI cases White men 17,902 130 16 White women 20,142 35 7 Black men 8,832 17 3 Black women 11,253 11 1 Total 58,129 193 27 Race‐specific MI mortality rate: ‐ MI mortality rate in blacks: = [(3+1)/(8,832+11,253)]*100,000 = 20 per 100,000 individuals 70 ‐ Also calculate rate for whites  More Practice Calculate the MI sex‐and‐race‐specific mortality rate: – For white women – For black men Calculate the prevalence of MI Calculate the sex ratio (male/female) for MI. prevalence among whites 69  Answers Calculate the MI sex‐and‐race‐specific mortality rate: – For white women= (7/20,142) – For black men= (3/8832) Calculate the prevalence of MI = (193/58,129)= 0.003 3 cases per 1000 population 72  Answers Calculate the sex ratio (male/female) for MI prevalence among whites – MI prevalence in white men= (130/17,902) – MI prevalence in white women= (35/20,142) – Sex ratio= (130/17,902)/(35/20,142)= 4.4 – Interpretation? 73 Sources of Morbidity and Mortality Data  Sources of Data Measures of morbidity and mortality can be obtained from different sources: 1.Conduct your own study to collect data on disease mortality and morbidity 2.Rely on available data (already collected for other purposes) – Vital statistics (death certificates), medical records, disease surveillance programs, registries, etc.. 73  Death Certificates Include information about: – Cause of death: Immediate cause: the condition or disease complication that directly preceded death (e.g. rupture of myocardium, minutes before death) Intervening cause: the conditions that caused the immediate cause and resulted from the underlying cause (e.g. acute myocardial infarction, 6 days before death) Underlying cause: the disease or injury which initiated the train of morbid events leading to death (e.g. chronic ischemic heart disease, 5 years before death) Associated cause – Demographic characteristics (age, sex, race) – Date and place of death 76 77  Death Certificates Limitations: – Specified cause of death may not be accurate; described in vague non‐specific terms – Lack of standardization in diagnostic criteria employed by different hospitals and physicians – Incomplete data (e.g. cause of death may not be listed for some diseases because of stigma) 78  Death Certificates Some reported causes of death, US : – “Died suddenly, nothing serious” – “Went to bed feeling well, but woke up dead” – “Deceased had never been fatally ill” – “Died suddenly without the aid of a physician” 79  Public Health Surveillance Programs Public health surveillance refers to the systematic and continuous gathering of information about the occurrence of diseases and other health phenomena. Monitors changes in disease frequency Originally conducted for infectious diseases, but now it is being increasingly used for chronic diseases and risk factors 80  Public Health Surveillance Programs (cont’d) E.g.: Reportable and Notifiable Disease Statistics: – It’s mandatory for physicians and health care providers to report certain diseases to health authorities (e.g. Epidemiologic Surveillance Unit of the MOPH in Lebanon ): Communicable diseases (rubella, tetanus, sexually transmitted disease, plague, food‐borne disease, etc..) http://www.moph.gov.lb/Prevention/Surveillance/Page s/Surveillance.aspx 81 Farhat G, Spring 2013 82 Public Health Surveillance Programs (cont’d) Limitations: – Incomplete coverage of population and under‐ reporting of cases: Not every person who develops the disease may seek medical care Failure of physicians to report on all cases Harder communication with rural areas 83  Other Sources of Data Disease registries: – Registry= centralized database for collection of information about a disease – E.g. National Cancer Registry in Lebanon Medical records National health surveys (on‐going and ad‐hoc) Birth statistics 84  Other Sources of Data Population census Needed to provide denominator for measures of morbidity & mortality. Population census is a complete snapshot of a nation's people. It provides information on the size, location, and characteristics of a population. It is the backbone of a national statistical system 85 Issues with Measures of Disease Occurrence Quality of data (completeness, representativeness) Problems with “numerator”: – Defining who has the disease Different diagnostic criteria used to define certain disease (e.g. dementia, rheumatoid arthritis) Changes in diagnostic criteria over time – Prevalence and incidence will vary depending on which diagnostic system is used Problems with “denominator” Are changes real or artifactual? 86