Inappropriate comparisons of incidence and prevalence in epidemiologic research

A science impact framework to measure impact beyond journal metrics

Measuring the impact of public health science or research is important especially when it comes to health outcomes. Achieving the desired health outcomes take time and may be influenced by several contributors, making attribution of credit to any one entity or effort problematic. Here we offer a science impact framework (SIF) for tracing and linking public health science to events and/or actions with recognized impact beyond journal metrics. The SIF was modeled on the Institute of Medicine's (IOM) Degrees of Impact Thermometer, but differs in that SIF is not incremental, not chronological, and has expanded scope. The SIF recognizes five domains of influence: disseminating science, creating awareness, catalyzing action, effecting change and shaping the future (scope differs from IOM). For public health, the goal is to achieve one or more specific health outcomes. What is unique about this framework is that the focus is not just on the projected impact or outcome but rather the effects that are occurring in real time with the recognition that the measurement field is complex, and it takes time for the ultimate outcome to occur. The SIF is flexible and can be tailored to measure the impact of any scientific effort: from complex initiatives to individual publications. The SIF may be used to measure impact prospectively of an ongoing or new body of work (e.g., research, guidelines and recommendations, or technology) and retrospectively of completed and disseminated work, through linking of events using indicators that are known and have been used for measuring impact. Additionally, linking events offers an approach to both tell our story and also acknowledge other players in the chain of events. The value added by science can easily be relayed to the scientific community, policy makers and the public.

A Meta-Analysis of the Association of Chronic Obstructive Pulmonary Disease with Abdominal Aortic Aneurysm Presence

BACKGROUND

Several case-control and population-based abdominal aortic aneurysm (AAA) screening studies have reported inconclusive results of the association of chronic obstructive pulmonary disease (COPD) with AAA presence. To determine whether COPD is associated with AAA presence, we performed a meta-analysis of contemporary clinical studies.

METHODS

To identify all contemporary case-control and population-based AAA screening studies evaluating the association of COPD with AAA presence, databases including MEDLINE and EMBASE were searched from January 2000 to May 2015 using Web-based search engines (PubMed and OVID). An adjusted odds ratio (OR) and 95% confidence intervals (CI) for COPD or AAA presence (using multivariable logistic regression) were abstracted from each individual study. We took an OR for AAA presence to be representative of an OR for COPD presence.

RESULTS

Of 159 potentially relevant articles screened initially, there were 7 case-control and 4 population-based AAA screening studies that met eligibility requirements and were included. Pooled analysis of all the 11 studies (14 estimates, 155,731 participants), 7 case-control studies (4171 participants), and 4 population-based AAA screening studies (7 estimates, 151,560 participants) respectively demonstrated a statistically significant 1.78-fold (OR 1.78, 95% CI 1.38-2.30, P < 0.00001), 3.05-fold (OR 3.05, 95% CI 1.44-6.49, P = 0.004), and 1.24-fold (OR 1.24, 95% CI 1.04-1.48, P = 0.02) increased prevalence/incidence of COPD in patients with AAA relative to subjects without AAA (i.e., a statistically significant 1.78-, 3.05-, and 1.24-fold increased prevalence/incidence of AAA in patients with COPD relative to subjects without COPD) (P for subgroup differences = 0.02).

CONCLUSION

The present meta-analysis demonstrated 1.8-fold increased prevalence/incidence of COPD in patients with AAA relative to subjects without AAA (i.e., 1.8-fold increased prevalence/incidence of AAA in patients with COPD relative to subjects without COPD), which suggests that COPD is associated with AAA presence.

Monitoring Clostridium difficile infection in an acute hospital: prevalence or incidence studies?

BACKGROUND

Surveillance of Clostridium difficile infection (CDI) is an essential component of a CDI preventative programme.

AIMS

The aim of this study was to evaluate two methods of CDI surveillance.

METHODS

Prevalence of CDI, antibiotic use and associated co-morbidity was assessed weekly on two wards over 6 weeks. In addition, CDI incidence surveillance was performed on all new CDI cases over a 13-week period. Cases were assessed for CDI risk factors, disease severity, response to treatment and outcome at 6 months.

RESULTS

Clostridium difficile infection prevalence was 3.5% (range 2.9-6.1%) on the medical ward and 1.1% (range 0-3.5%) on the surgical ward. Patients on the medical ward were older and more likely to be colonised with MRSA; however, recent antibiotic use was more prevalent among surgical patients. Sixty-one new CDI cases were audited. Patients were elderly (mean age 71 years) with significant co-morbidity (median age adjusted Charlson co-morbidity score 5). CDI ribotypes included 027 (29 cases) 078 (5) and 106 (4). Eight patients developed severe CDI, seven due to 027. Antibiotic use was common with 56% receiving three or more antibiotics in the preceding 8 weeks. Twenty-four patients had died at 6 months, five due to CDI.

CONCLUSION

Clostridium difficile infection prevalence gives a broad overview of CDI and points to areas that require more detailed surveillance and requires little time. However, patient-based CDI incidence surveillance provides a more useful analysis of CDI risk factors, disease and outcome for planning preventative programmes and focusing antibiotic stewardship efforts.

Estimating prevalence in single-gene kidney diseases progressing to renal failure

Incidence and prevalence, the measures of "frequency, " are often confused. While in a nonhereditary situation, the useful parameter is the incidence rate, evaluating the impact of an etiologic factor, it is prevalence that is considered useful in a hereditary disease. Prevalence may concern either the whole population or a fraction of this population, that is, males or females or individuals at a given age, for example, at birth. Pathologic phenotype and morbid genotype prevalences have to be clearly differentiated. In this article, we review the epidemiologic surveys allowing an estimation of the distribution of major single-gene kidney diseases progressing to renal failure in different populations. In order to compare their results, the geographic/ethnic composition of the population, the determination of its size, the choice and mode of calculation of the epidemiologic measure, the definition of the disease and modes of diagnosis, the inclusion of cases, the sources of ascertainment and the possible causes of underascertainment, and the period of time during which events were counted should be analyzed accurately. Although their impact in terms of morbidity, hospitalizations, mortality, and cost to society is high, this review shows that information on the prevalence of single-gene kidney diseases is far from complete. To date, the data essentially apply to large populations of European origin. A part of the variation among prevalence data may be due to methodological differences. Not representative are the small populations in which some rare diseases, especially recessive, are found with a high prevalence.

On the terminology and dimensions of incidence

This paper is a review of the concept of incidence and the measures of incidence commonly used. The three components of incidence (observed event, population at risk, and observation period) and the four measures of incidence (incidence as number of events, as number of events per time unit, as number of events per unit of amounts of observation, and as probability) are discussed. The terminology and dimensions of incidence measures used in epidemiology literature are reviewed. Ambiguities and uses of the same term for different purposes seem to suggest the need for standardized terminology. Incidence measures with length-based amount of observation are also briefly discussed.

Nosocomial Infections in Spain: Results of Five Nationwide Serial Prevalence Surveys (EPINE Project, 1990–1994)

Objective: To determine trends in rates of nosocomial infections in Spanish hospitals.

Design: Prospective prevalence studies, performed yearly from 1990 through 1994.

Setting: A convenience sample of acute-care Spanish hospitals.

Participants and Patients: The number of hospitals and patients included were as follows: 1990, 125 hospitals and 38,489 patients; 1991, 136 and 42,185; 1992, 163 and 44,343; 1993, 171 and 46,983; 1994, 186 and 49,689. A core sample of 74 hospitals, which participated in all five surveys and included a mean of 23,871 patients per year, was analyzed separately.

Results: The overall prevalence rate of patients with nosocomial infections in the five studies was as follows: 1990, 8.5%; 1991, 7.8%; 1992, 7.3%; 1993, 7.1%; and 1994, 7.2%. The prevalence rate of patients with nosocomial infection in the core sample of 74 hospitals was 8.9%, 8.0%, 7.4%, 7.6%, and 7.6%, respectively (test for trend, P=.0001). Patients admitted to intensive care units had a 22.8% prevalence rate of nosocomial infection in 1994. The most common nosocomial infections by primary site were urinary tract infection and surgical site infections, followed by respiratory tract infections and bacteremia. More than 60% of all infections were supported by a microbiological diagnosis.

Conclusions: The EPINE project provides a uniform tool for performing limited surveillance of nosocomial infections in most Spanish acute-care hospitals. Its use helps to spread an accepted set of definitions and methods for nosocomial infection control in the Spanish healthcare system. The surveys indicate that the prevalence of nosocomial infections has been reduced over the last 5 years in a core sample of Spanish hospitals.

Comparing period prevalences with application to drug utilization

Period prevalence is frequently measured in studies based on administrative data such as that from health maintenance organizations. For example, treated prevalence and drug utilization prevalence are important measures that are typically defined in relationship to a specified time period. Often one wishes to compare administrative data with period prevalences based on national surveys. It may also be of interest to compare period prevalences from two (or more) different data sources. This comparison is not straightforward owing to the problem of "person-time at risk." This article reviews the values and drawbacks of period prevalence as compared with cumulative incidence. L ife table methodology is described for comparing period prevalence data from administrative databases with survey results. This technique can be extended to the comparison of period prevalence observations from two or more administrative data bases. Examples are given pertaining to hypnotic drug use and the treatment of schizophrenia.

A mathematic analysis of the "epidemiologic necropsy"

There has been much debate regarding the validity and implications of the "epidemiologic necropsy." Two opposing mathematic analyses have previously appeared--one by Flanders and O'Brien and one by Feinstein and colleagues. Because these analyses omitted discussion of certain key parameters and assumptions, I provide a more detailed analysis. I show that under the assumptions used by earlier authors, the observed ("registry") disease rates will equal the true disease rates, even if there is differential disease detection and large pool of undetected cases. Under realistic violations of assumptions, both the observed rates and the necropsy estimates will be biased. In any case, the data offered thus far are insufficient to accurately assess the impact of undetected disease on epidemiologic studies.