Evaluation of program performance and expenditures in a report of performance measures (RPM) via a case study of two Florida county tuberculosis programs.

Health Department (HD) managers at both state and local levels are in desperate need of tools to assist in monitoring and evaluating programs. The purpose of this study is to assess the feasibility and utility of linking program performance scores and expenditures into a Report of Performance Measures (RPM). We analyzed secondary data on performance indicators, selected by HD staff, and expenditures, related to six surveillance activities, from two, similar, high-incidence, tuberculosis (TB) programs in Florida from 2002 to 2003. We compared the findings between the county HDs as an illustration of basic cost-effectiveness benchmarking, based on the cost-effectiveness grid. Data included here provide examples of: (1) two instances in which one county was operating relatively inefficiently compared to the other; (2) two instances in which performance and expenditures were similar for the counties; and (3) two instances in which one county spent more for higher performance scores than the other. These data illustrate how the RPM can be used to facilitate benchmarking, a basic evaluation tool. They also demonstrate ways to identify potential operational inefficiencies in a single time period and ultimately over time. It is thus likely to be a feasible and useful management tool.

Establishment of public health security in Saudi Arabia for the 2009 Hajj in response to pandemic influenza A H1N1.

Mass gatherings of people challenge public health capacities at host locations and the visitors' places of origin. Hajj--the yearly pilgrimage by Muslims to Saudi Arabia--is one of the largest, most culturally and geographically diverse mass gatherings in the world. With the 2009 pandemic influenza A H1N1 and upcoming Hajj, the Saudi Arabian Ministry of Health (MoH) convened a preparedness consultation in June, 2009. Consultants from global public health agencies met in their official capacities with their Saudi Arabian counterparts. The MoH aimed to pool and share public health knowledge about mass gatherings, and review the country's preparedness plans, focusing on the prevention and control of pandemic influenza. This process resulted in several practical recommendations, many to be put into practice before the start of Hajj and the rest during Hajj. These preparedness plans should ensure the optimum provision of health services for pilgrims to Saudi Arabia, and minimum disease transmission on their return home. Review of the implementation of these recommendations and their effect will not only inform future mass gatherings in Saudi Arabia, but will also strengthen preparedness efforts in other settings.

Epidemic investigation of diphtheria in the Republic of Armenia, 1990-1996.

While incidence rates of diphtheria steadily declined in Armenia after World War II, reemergence of the disease in 1990 brought about changes in public health practices and identified resource needs. The Armenian Ministry of Health (MOH) routinely collected diphtheria case reports, as a reportable health outcome. Diphtheria incidence rates increased from 0.02/100,000 in 1993 to 1/100,000 (36 cases) in 1994. The distribution of cases showed that the greatest number of illnesses and deaths occurred among persons 5-14 years old, yet incidence rates among persons 1-4 and 5-14 years old were similar (4. 4 cases/ and 4.3 cases/100,000, respectively). During 1990-1996, 9 (75%) of 12 cases who died and 18 (21%) of 84 cases who survived had not been vaccinated. The diphtheria epidemic in Armenia was an important, serious, and signal public health event. The Armenian MOH responded by revising immunization practices (1994), improving epidemic control measures (1995), and soliciting international resources (1992-1996).

Epidemic investigation of diphtheria, Republic of Kazakhstan, 1990-1996.

The diphtheria epidemic that began in Russia in 1990 reached Kazakhstan in 1992 when 45 case-patients (a 50% increase over 1991) were reported. In 1993, 82 case-patients were reported, and 489 were reported in 1994. The epidemic peaked in 1995 when 1105 case-patients were reported (incidence rate=6.7/100,000 population). In 1996, after public health practice modifications and several mass vaccinations, 455 case-patients were reported. From 1990 to 1996, children

Deaths related to Hurricane Andrew in Florida and Louisiana, 1992.

BACKGROUND: Information about circumstances leading to disaster-related deaths helps emergency response coordinators and other public health officials respond to the needs of disaster victims and develop policies for reducing the mortality and morbidity of future disasters. In this paper, we describe the decedent population, circumstances of death, and population-based mortality rates related to Hurricane Andrew, and propose recommendations for evaluating and reducing the public health impact of natural disasters. METHODS: To ascertain the number and circumstances of deaths attributed to Hurricane Andrew in Florida and Louisiana, we contacted medical examiners in 11 Florida counties and coroners in 36 Louisiana parishes. RESULTS: In Florida medical examiners attributed 44 deaths to the hurricane. The mortality rate for directly-related deaths was 4.4 per 1 000 000 population and that for indirectly-related deaths was 8.5 per 1 000 000 population. In Louisiana, coroners attributed 11 resident deaths to the hurricane. Mortality rates were 0.6 per 1000 000 population for deaths directly related to the storm and 2.8 for deaths indirectly related to the storm. Six additional deaths occurred among non-residents who drowned in international waters in the Gulf of Mexico. In both Florida and Louisiana, mortality rates generally increased with age and were higher among whites and males. CONCLUSIONS: In addition to encouraging people to follow existing recommendations, we recommend emphasizing safe driving practices during evacuation and clean-up, equipping shelters with basic medical needs for the population served, and modifying zoning and housing legislation. We also recommend developing and using a standard definition for disaster-related deaths, and using population-based statistics to describe the public health effectiveness of policies intended to reduce disaster-related mortality.

Correlates of gun-carrying among adolescents in south Louisiana.

In the majority of episodes of fatal interpersonal violence, the weapon used is a firearm. Amid frequent reports of youths carrying weapons, including firearms, we conducted a case-control study to identify risk factors for being charged with gun-carrying and gun-carrying, per se, among adolescents in Jefferson Parish, Louisiana. Cases were defined as incidents of gun-carrying among adolescents < 19 years of age, legally charged in the Jefferson Parish, Louisiana, juvenile court from January 1, 1992, through April 15, 1993. For each case, we randomly drew three age-, gender-, and school-matched control subjects from the enrollment rosters of the Jefferson Parish public schools and administered a questionnaire. The data set comprised 38 case subjects and 103 matched control subjects. Thirty (29%) control subjects reported gun-carrying. Both case subjects and gun-carrying control subjects reported self-defense (40%) as the main reason for gun-carrying. Most case subjects (25 [66%]) were African Americans, but only 8 (27%) gun-carrying and 27 (37%) non-gun-carrying control subjects were African Americans. Case subjects were significantly more likely than gun-carrying control subjects to report being African American (odds ratio [OR] = 5.3, 95% confidence intervals [CI] = 1.6, 17.5). In crude analyses, case subjects were more likely than non-gun-carrying control subjects to report adult-male unemployment among households with adult men, to foresee a likelihood to be shot in school, to have seen a shooting, to use marijuana, to watch television > 6 hours per day, and to be African American. After the effect estimates were adjusted in conditional logistic regression modeling, case subjects were more likely than non-gun-carrying control subjects to report adult-male unemployment among households with adult men, using marijuana, and watching television > 6 hours per day (OR = 8.6, 95% CI = 1.2, 61,2; OR = 11.7, 95% CI = 2, 70.2; and OR = 6.5, 95% CI = 0.8, 51.9, respectively). Gun-carrying control subjects were significantly more likely than non-gun-carrying control subjects to report their school not safe, having seen a shooting, using marijuana, and having fired a gun (OR = 9, 95% CI = 1, 82.1; OR = 7, 95% CI = 1.3, 38.2; OR = 6.8, 95% CI = 1.8, 25.5; and OR = 17, 95% CI = 1.8, 156.6, respectively). We found that gun-carrying was very common, and that adolescent youths who carry guns were more likely to have familiarity with guns and experience with or perception of an unsafe environment. Together, these lead to the conclusion that gun-carrying is a common response of youths who live in a risky environment, who do not have the social support to learn how to deal effectively with that risk, and who have access to guns, which they think may provide them with some protection.

Hurricane Andrew-related injuries and illnesses, Louisiana, 1992.

To determine the extent and types of injuries and illnesses in Louisiana associated with or related to Hurricane Andrew, we gathered data from hospital emergency departments and coroner's offices on demographic variables, institution, nature and cause of the injury or illness, body part affected, location, and date and time of the event. A hurricane-related injury or illness was defined as one that occurred from noon on August 24, 1992, through midnight on September 21, 1992, as a direct or indirect result of the preparation for (preimpact), the impact of, or the clean-up after the hurricane (postimpact). Nineteen parishes in south-central Louisiana that were most affected by Hurricane Andrew provided data from patients seen in emergency departments and reports from coroner's offices. Active, advance surveillance of this type promotes and facilitates the reporting of disaster-related health outcomes. Future planning for hurricanes should take into account the high rate of cuts, lacerations, and puncture wounds, particularly during the postimpact phase.

Injuries to international petroleum drilling workers, 1988 to 1990.

Nonfatal work-related injury (NFI) rates are 49% higher among oil and gas field workers than among workers in all US industries combined, and these injuries are more severe (the rate of lost workdays in the oil and gas field services industry is 2.8 times that of all US industries combined). We analyzed the 1988 to 1990 incident reports submitted by drilling companies to the International Association of Drilling Contractors, an industry-wide international trade association representing 95% of the world's oil and gas drilling companies. We determined geographic and occupation-specific incidence rates by full-time equivalents, calculated per job category and year. Of the 5,251 reports, 5,218 (99.4%) were of NFI and 33 (0.6%) of fatal work-related injuries (FI). The overall NFI rate was 1.2/100 full-time equivalents and the overall FI rate was 7.5/100,000 full-time equivalents. Reported NFI in US territory was 4 times more common than in non-US territory. Reported FI in US and non-US waters were 4 and 5 times more common than on land, respectively. Three job categories--floormen, roustabouts, and derrickmen--accounted for 74% of the NFI and 64% of FI, with a rate ratio, compared with rates for all other occupations, of 10.5, 8.5, and 7.0 for NFI and 5.0, 9.4, and 4.0 for FI. Among all occupations, the body part most frequently injured was the upper extremity (1,631/5,218 [31%]). The four key NFI types and circumstances identified included the upper extremities "caught in" (857/5,218 [16%]),the back "strained" (592/5,218[11%]), the lower extremities "struck by" (538/5,218 [10%]), and the lower extremities injured while "slipping" (402/5,218 [8%]). Results of these analyses revealed several high-risk occupations in this industry and identified high-risk activities that can be targeted for further study.

Population-based nutritional risk survey of pensioners in Yerevan, Armenia.

Armenia, a republic of the former Soviet Union, currently suffers from hyperinflation of its currency, a five-year country-wide blockade, and a war with Azerbaijan. Pensioners 60 years of age or older may be at high risk for significant nutritional deficits. We drew a stratified systematic sample (with a random starting point) of 456 pensioner names from all eight administrative regions in Yerevan, the capital of Armenia. We administered a questionnaire that gathered data including self-reported weight and height, demographic characteristics, living conditions, medical and dietary history, income, and aid received from various sources. The survey yielded 381 of 456 (84%) completed interviews. Ninety-one percent reported their diet had gotten worse during the past six months, including less variety (83%) and quantity (85%) of food. Seventy-six percent reported they did not have enough money to buy food, and 91% had cut the size of their meals or skipped meals. Forty-five percent reported a weight loss of > or = 5 kg in the previous year. After we adjusted for potential confounders, weight loss of > or = 5 kg was associated with illness affecting eating (adjusted odds ratio [OR] = 2.2, 95% confidence intervals [CI] = 1.4, 3.4), not having received aid (adjusted OR = 2.2, 95% CI = 1.1, 4.1), and cutting the size of or skipping meals (OR = 2.7, 95% CI = 1.1, 6.7).(ABSTRACT TRUNCATED AT 250 WORDS)

Tularemia pneumonia in Oklahoma, 1982-1987.

OBJECTIVE: We assessed the incidence, risk factors, and prognostic implications of tularemia pneumonia in Oklahoma from 1982 through 1987. DESIGN: We retrospectively reviewed all reported case-patients over the six-year period 1982-1987. SETTING: Department of Health in Oklahoma, where tularemia is known to be endemic. PATIENTS: Of 128 patients with tularemia who entered the study, 32 had pulmonary involvement. MAIN RESULTS: Compared to patients without pulmonary involvement, those with tularemia pneumonia were older (52 vs 32, p less than .0001), less likely to give a history of vector exposure (25% vs 7%, p less than .05), more likely to present with typhoidal illness (56% vs 15%, p less than .0001), hospitalized longer (11.6 vs 4.7 days, p less than .001), more likely to have a positive culture (9 vs 7, p less than .01), and more likely to die (4 vs 1, p less than .01). CONCLUSIONS: Patients with tularemia pneumonia often present without historical or physical examination findings that suggest the diagnosis; thus, tularemia pneumonia often cannot be distinguished from other cases of community-acquired pneumonia. Therefore, especially in areas where the disease is endemic, tularemia must be considered in patients with pneumonia.

The descriptive epidemiology of Lyme disease in Oklahoma.

Lyme disease is a tick-borne illness that primarily occurs in the United States in three endemic areas: the northeast, upper midwest, and pacific coastal regions. Although Oklahoma is considered a non-endemic area of Lyme disease, other tick-borne infections such as Rocky Mountain spotted fever, ehrlichiosis, and tularemia occur at endemic levels. In order to determine the extent of Lyme disease in Oklahoma, the Oklahoma State Department of Health collected information on all possible cases of Lyme disease. The first reported Oklahoma case occurred in 1985. In 1988, and 1989, 8 and 25 cases, respectively, of Lyme disease were reported in Oklahoma residents who acquired their infection indigenously. The mean age of case-patients was 38 years. Twenty-eight of 33 (85%) case-patients were white, and 4 of 33 (12%) were American Indian (race was unknown for 1 case-patient). The female/male ratio was 2.3. Most case-patients had onset of illness between May and September. Twenty-two counties reported cases, with Oklahoma County accounting for 4 of 33 (12%). Only 21 of 33 (64%) recalled a tick bite, and only 19 of 33 (58%) had erythema migrans. Continued active case-finding and passive reporting (as now mandated by state law) will further increase our knowledge of the epidemiology, ecology, and prevention of Lyme disease in Oklahoma.

Pneumocystis carinii antigen detection in rat serum and lung lavage.

We developed a modified double-antibody sandwich enzyme-linked immunosorbent assay (ELISA) that detected relatively low concentrations of known Pneumocystis carinii antigen added to buffer or rat sera. Artificial immunization-derived polyclonal rabbit anti-P. carinii antibody was used on the solid phase to capture the antigen. Infection-derived (after P. carinii pneumonia) polyclonal rat anti-P. carinii antibody or a mixture of five murine monoclonal antibodies was used as the antigen detector antibody. Rabbit anti-rat immunoglobulin G antibody or goat anti-mouse immunoglobulin G antibody conjugated to alkaline phosphatase was used as the final antibody. After standardization and optimization of the various reactants in this ELISA system, approximately 53 ng of known P. carinii antigen per ml suspended in phosphate-buffered saline-Tween 20 buffer or 210 ng of antigen per ml suspended in normal rat serum diluted 1:4 could be detected. In addition, an indirect ELISA for P. carinii antibody measurement was developed, using as the antigen a soluble supernatant from a sonicated preparation of Percoll-purified whole cysts and trophozoites to coat the solid phase. Limited studies with sera from a small number of caesarian-obtained, barrier-sustained rats from Charles River Breeding Laboratories, Inc., and the National Institutes of Health and sera from normal and heavily infected rats indicated that the caesarian-obtained, barrier-sustained rats had negligible levels of antibody. The normal and heavily infected rats had variable antibody titers. A significantly high level of P. carinii antigenemia was detected in only 2 (11%) of 18 heavily infected rats. Extensive studies of the P. carinii pneumonia rat model with the ELISA did not reveal significant serum P. carinii antigenemia during the acute stage of infection. However, soluble P. carinii antigen was detected by the ELISA and Western blot assays in the supernatant of lavage fluid after centrifugation to sediment intact organisms. As expected, P. carinii antigens were detected by these assays in the lavage pellet recovered after centrifugation. In conclusion, the antigen assay used in this study detected P. carinii antigen in lung lavage but failed to detect P. carinii antigen in rat serum during the acute phase of infection.

Outbreak of diarrhea in a day care center with spread to household members: the role of Cryptosporidium.

From August 1 to November 1, 1984, 80 (43%) of 186 children attending a day care center (DCC) in Tulsa, OK, had diarrhea. Seventy-one percent of children less than 3 years of age had diarrhea, compared with 17% of children 3 to 8 years old. The duration of illness was 1 to 42 days (median, 8 days). At least one stool specimen was obtained from 142 (76%) of the 186 children. Children who had had diarrhea were more likely to have a specimen positive for Cryptosporidium (27 of 77, 35%) than were children who had not had diarrhea (8 of 65, 12%) (P less than 0.01). The presence of Giardia, Campylobacter, enterovirus, or rotavirus was not statistically associated with diarrhea. These results suggest Cryptosporidium as a causative agent. The secondary attack rate for diarrhea among household contacts exposed to DCC children with diarrhea was 77 of 204 (38%) compared with 25 of 273 (9%) for household contacts of children in the DCC who did not have diarrhea (P less than 0.001). Stool specimens were obtained from 79 household contacts. Six (23%) of 26 household members exposed to a Cryptosporidium-positive child had a stool specimen positive for Cryptosporidium, compared with 1 (2%) of 53 household contacts exposed to a DCC child whose stool specimen was negative for Cryptosporidium (P less than 0.01). The outbreak resolved within 2 weeks after cohorting children with diarrhea and stressing the importance of handwashing among staff and older children. cryptosporidium can cause outbreaks of diarrhea in DCC settings, can be spread within households and may be more common than previously recognized among DCC attendees and their household members.

Analyses of rat Pneumocystis carinii antigens recognized by human and rat antibodies by using western immunoblotting.

The major Pneumocystis carinii antigens inducing antibody responses in infected hosts were identified by Western immunoblotting techniques. The biochemical nature of these antigens was also elucidated. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis followed by protein staining revealed a major component with a molecular weight (MW) of greater than 205,000. This major component disappeared and a new major protein staining component of approximately 110,000 to 116,000 MW appeared when electrophoresis was done in the presence of beta-mercaptoethanol. Periodic acid-Schiff staining revealed that this major component contains carbohydrate moieties. A major component in the 55,000- to 60,000-MW region was visible with periodic acid-Schiff stain, but not with a protein stain, after electrophoresis in the presence of beta-mercaptoethanol. The majority of sera tested from humans with diagnosed pneumocystosis and from rats allowed to recover from steroid-induced pneumocystosis reacted strongly with 110,000- to 116,000-, and 55,000- to 60,000-MW components. These sera often, but not always, detected antigens with MWs of approximately 170,000, 125,000, and 30,000 to 32,000. The data suggest that the antigenic composition of P. carinii is relatively complex and that rat and human P. carinii probably share antigenic determinants. Competitive studies between infection-derived human and rat antisera for the major rat P. carinii components revealed competition; rat antisera appeared to recognize a greater range of antigenic epitopes than did human antisera. Protease treatment of the antigenic components that had been immobilized on nitrocellulose paper destroyed their antigenic reactivity with rat antibody. Treatment with sodium periodate decreased reactivity of this 110,000- to 116,000-MW component and completely destroyed the reactivity of the 55,000- to 60,000-MW component with rat antibody.

Development and characterization of monoclonal antibodies to Pneumocystis carinii.

Hybridoma-producing monoclonal antibodies against Pneumocystis carinii were produced by the fusion of nonsecreting mouse myeloma cells (P3X63-Ag8.653) with splenocytes from BALB/c mice that had been immunized with partially purified preparations of P. carinii. Of 227 hybridoma clones producing antibodies against P. carinii, as measured by an enzyme-linked immunosorbent assay, 12 monoclonal antibodies showing the highest reactivity in the enzyme-linked immunosorbent assay were further characterized. The majority (11 of 12) of the monoclonal antibodies did not cross-react with Candida albicans, Cryptococcus neoformans, Histoplasma capsulatum, or Mycobacterium avium as determined by absorption experiments. By using the indirect immunofluorescence assay, serological reactivity was shown for these antibodies with titers ranging from 1:40 to 1:10,240. By using a competitive binding assay, these 12 monoclonal antibodies could be divided into seven groups, each group reacting with a different antigenic determinant of P. carinii. Sodium dodecyl sulfate-polyacrylamide gradient gel electrophoresis of P. carinii, followed by Western immunoblot analysis, allowed the identification of one major antigen with an apparent molecular weight of 110,000 by all 12 monoclonal antibodies. Other minor bands with molecular weights of approximately 116,000, 90,000, 55,000, and 35,000 were recognized by several of the monoclonal antibodies.

Comparison of sedimentation and flotation techniques for identification of Cryptosporidium sp. oocysts in a large outbreak of human diarrhea.

Cryptosporidiosis, previously seen mostly among immunocompromised patients, is now recognized among immunocompetent patients. During a large outbreak of cryptosporidiosis in two day-care centers, we compared two procedures for the demonstration of the organism in preserved stool specimens. Of 703 stool specimens tested by both techniques, Sheather sucrose flotation (SSF) identified 127 (18.1%) as positive for Cryptosporidium sp. oocysts. Ritchie Formalin-ethyl acetate sedimentation (F/EA) plus a modified cold Kinyoun acid-fast stain (MCK) of the sediment identified 129 (18.4%) as positive for Cryptosporidium sp. oocysts. The degree of agreement between the two tests was statistically highly significant (P less than 0.0001). A total of 161 (22.9%) were positive by one technique or the other; 95 (13.5%) were positive by both techniques. A total of 32 specimens were positive by SSF but negative by F/EA plus MCK, and 34 specimens were positive by F/EA plus MCK but negative by SSF. The discrepancies between the two techniques occurred in stool specimens that contained rare to a few oocysts. Other parasitic forms were found by both techniques. F/EA plus trichrome staining recovered 126 (17.9%) specimens with Giardia lamblia, whereas SSF recovered only 42 (6.0%) specimens with G. lamblia. No association (chi 2 = 0.02, P = 0.89) was observed between the presence of G. lamblia and Cryptosporidium sp. in these stool specimens. We concluded that F/EA plus MCK of the sediment was as effective in the concentration and identification of Cryptosporidium sp. oocysts as SSF. F/EA plus MCK may be advantageous as a single concentration method for general parasitology when Cryptosporidium sp. is also being sought.

Rapid trichrome stain.

A rapid trichrome stain procedure was evaluated. It saves considerable time, especially when only a few slides are stained, and allows for the accurate identification of amoebae and other fecal parasites.