The Michigan Disease Surveillance System for histoplasmosis reporting was evaluated according to CDC guidelines. Several key findings were: 1) The system was found to be stable, flexible, and relatively simple to use, though the sensitivity was low at 27.09% and misclassification of cases was high. 2) Reporting timeliness improved after a 2007 case definition implementation, with investigation completion times decreasing from 48 to 31 days on average. 3) The system was fairly representative though underrepresented African Americans; recommendations were made to improve case classification and data quality.

1. Evaluation of the Michigan Disease Surveillance System for Histoplasmosis Reporting
Veronica A. Fialkowski, MPH; Leigh Tyndall Snow, MPH; Kimberly Signs, DVM; Mary Grace Stobierski, DVM, MPH
Michigan Department of Health and Human Services
Histoplasmosis is an infectious disease caused
by a fungus called Histoplasma capsulatum.
Fungal spores are found in the soil. These spores
are mostly associated with bird and bat droppings,
and if inhaled can cause lung infection.
The disease is endemic in the
Ohio and Mississippi River Valley.
Michigan reports >100 cases/year.
Histoplasmosis is a reportable disease in Michigan
to identify increase in incidence and potential
point-source outbreaks. A case definition was
implemented in 2007.
Cases are reported into the Michigan Disease
Surveillance System (MDSS), a web-based
electronic database, and are investigated by local
health departments (LHD).
Attribute Analysis Methods
Representativeness: demographic and geographic
characteristics of the cases were compared to the
Michigan population
Acceptability: matched a comprehensive list of
hospitals in Michigan to hospitals that directly
report into MDSS
Sensitivity: two-source capture-recapture method
using the Michigan Inpatient Database as the gold
standard to match hospitalized cases in MDSS by
date of birth and zip code
Positive predictive value (PPV): cases from 2014
were reviewed to determine if cases met the case
definition criteria and PPV was then calculated for
misclassified cases in MDSS
Data Quality: percent of missing and unknown
variables in MDSS by case report form category
Simplicity: chart describing the flow of data from
initial diagnosis to a completed case in MDSS was
prepared; analyzed number of cases reported by
case status
Stability: considered number of reported
system unscheduled outages
Timeliness: average times between diagnosis date,
date reported to LHD, and investigation completion
date were calculated.
Flexibility: reviewed MDSS process for new
demand and changes
Conclusions
Overall the system is stable, flexible, accepted,
and relatively simple. With increased use of
electronic laboratory records, the advantage is
an increase in number of reports, but the
number of ‘not a case’ increases as well, which
increases LHD workload.
Michigan’s histoplasmosis surveillance system
is fairly representative, however African
Americans are under-represented. This is most
likely an artifact of healthcare access or
geographical differences in disease endemicitiy
(rural vs. urban).
The misclassification of cases is troublesome,
with many cases not appropriately classified by
the Michigan case definition.
The sensitivity is alarming however, there are
several limiting assumptions made for analysis,
therefore the low sensitivity could be product
of these inherent limitations.
ACCEPTABILITY
TIMELINESS
SENSITIVITY
POSITIVE PREDICTED VALUE
SIMPLICITY
FLEXIBILITY
STABILITY
Prior to 2007, case investigation time averaged 48 days,
and decreased to a mean of 31 days after the
implementation of the case definition in 2007.
Diagnosis
Casereported
Investigation
completed
avg. 14 days
median=6 days
n=729
avg. 35 days
median=16 days
n=1145
Number of days to report and complete cases in
MDSS from 2004 to 2014:
PPV=79.84% (95% CI: 71.7%-86.5%)
MDSS is designed to allow for new demand. Changes to the
form can be made fairly easily, but do require funds and
personnel to adapt changes into the system.
Sensitivity=27.09%
(95% CI: 21.11-33.76%)
This study/report was supported in part by an appointment to the
Applied Epidemiology Fellowship Program administered by the
Council of State and Territorial Epidemiologists (CSTE) and funded
by the Centers for Disease Control and Prevention (CDC)
Cooperative Agreement Number 1U38OT000143-02
“Gold Standard”
Michigan Inpatient
Database (2012)
+ -
2012
MDSS
reported
cases*
+ 55 169 224
- 148
203
The histoplasmosis surveillance system was
evaluated according to CDC guidelines. Evaluating
a surveillance system is important to ensure that
problems of public health importance are
monitored effectively. This is the first evaluation of
the Michigan histoplasmosis surveillance system.
DATA QUALITY
0
5
10
15
20
25
<1 1-9 10-19 20-29 30-39 40-49 50-59 60-69 70-79 80+
Percent
Age category (years)
Histoplasmosis cases
Michigan demogrpahics
0
50
100
150
200
250
300
350
400
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014
Casecount
Year
Confirmed Not a Case Probable Total No. of Cases
In 2014, 72% of
histoplasmosis
cases reported
into MDSS did
not meet the
case definition.
Case definition
implemented
Increase in electronic
lab reporting
0
10
20
30
40
50
60
70
80
90
100
Caucasian African
American
Asian
Percent
Race
Histoplasmosis cases
Michigan demographics
62%
Male
38%
Female
REPRESENTATIVENESS
Recommendations to improve
histoplasmosis surveillance
Increasing the number of hospitals and
laboratories that report directly to MDSS and
train additional personnel to manage caseload.
Increase LHD awareness and distribution of
resources provided to aid in case classification.
Educate staff on laboratory results and
interpretations.
Emphasize the importance of case report form
variable completion for epidemiological
analysis.
Review case report form routinely.
Continue to increase physician – local health
department communication.
100% LHD buy-in since 2004.
58% of hospitals report directly to MDSS.
*includes all cases of
reported histoplasmosis
into MDSS in 2012 with
patient hospitalization
status marked as ‘yes’,
‘unknown’ or missing
Proportion of cases classified correctly that meet the
surveillance case definition for acute histoplasmosis in 2014.
MDSS is fully operating 99% of the time.
The number of unscheduled outages reported is
low. 2014 experienced a relatively high number
(approximately thirty) hardware issues and
unscheduled outages.
0
10
20
30
40
50
60
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014
Percentmissingdata
Year
Demographic
Clinical Information
Laboratory Information
Epidemiology
Person is exposed to
Histoplasma, 3-17 days
later feels ill, and seeks
medical attention
Hospital
Primary
Care
Physician
Laboratory Local
Health
Department
MI Disease
Surveillance
System
MI Dept. of Health and Human Services
or
Percent of missing data by case report form
category by year.
The process of histoplasmosis reporting.
The red arrows indicate the simplest flow of
information, the yellow arrows show the
process in practice.
Number of reports entered into MDSS by case status by year.
Race distribution of
histoplasmosis cases and
Michigan population.
African Americans are
under-represented.
Age distribution of histoplasmosis cases and Michigan
population. Ages 30-79 years are at highest risk for the disease.
≤10 per 100,00
10.1-30 per 100,00
≥55.1- per 100,00
30.1-55 per 100,00
Incidence of histoplasmosis
cases per 100,000 persons
in Michigan.
References
1. German RR, Lee LM, Horan JM, et al. Updated guidelines for evaluating
public health surveillance systems: recommendations from the Guidelines
Working Group. MMWR Recomm. Rep. 2001; 50 (RR-13):1-35.
2. Lenhart SW, Schafer MP, Singal M, et al. Histoplasmosis: Protecting
Workers at Risk. DHHS (NIOSH). 2004; 2005-109: 1-26
3. Wheat LJ, Freifeld AG, Kleiman MB, et al. Clinical Practice Guidelines for
the Management of Patients with Histoplasmosis: 2007 Update by the
Infectious Diseases Society of America. Clin. Infect. Dis. 2007; 45: 807-25.
4. Whitfield K, Kelly H. Using the two-source capture-recapture method to
estimate the incidence of acute flaccid paralysis in Victoria, Australia.
World Health Organization. 2002; 80: 846-851.
Acknowledgement:
MDHHS Bureau of Disease Control, Prevention, & Epidemiology Staff:
Edward Hartwick, MS, Tiffany Henderson, MPH
2004-2014 cases
Limitations:
1) Use of histoplasmosis ICD-9
codes for past infections in MIDB
2) Misclassification of the
hospitalization variable in MDSS