Descriptive Epidemiology and Analysis of Risk Factors for Invasive Disease Due to Serotype 1 Streptococcus pneumoniae in South Africa from 2003-2008 Open Access

Landrith, Tyler (2012)

Permanent URL:



Streptococcus pneumoniae is responsible for a variety of human infections, and is
a significant cause of morbidity and mortality among infants and the elderly, particularly
in developing countries. The pneumococcus has a total of 93 known immunologically
distinct polysaccharide capsules known as serotypes. Each serotype, including serotype 1
can exhibit unique epidemiological patterns based on factors such as invasiveness, age,
outbreak potential, disease outcome, and HIV infection. This study sought to elucidate
these patterns during the period of interest using three different epidemiological methods.
A description of cases and putative outbreaks of serotype 1 was conducted
according to time and place during the period 2003-2008 using data from a laboratory-
based surveillance system. Bivariate case-control analyses were used to test for
significant associations between serotype 1 and all relevant literature-based hypothesized
exposures, including HIV infection, relative to other serotypes causing invasive
pneumococcal disease. Bivariate analyses were also used to test for significant
associations between serotype 1 IPD and clinically relevant outcomes. A multivariable
model was created for refined risk factors for serotype 1 IPD using logistic regression.
Several multivariable models were created to test the significance of serotype 1 IPD as a
predictor of clinical outcomes, controlling for relevant covariates.

Serotype 1 was the most frequently isolated type in reported cases of IPD over the
period of interest (n=2701). The seven most frequently isolated serotypes over the period
of interest are all contained in the PCV13 vaccine, which is currently being rolled out in
South Africa. In Gauteng province, putative outbreaks occurred during July 2003 and
October 2008. Gauteng province also exhibited apparent seasonal variation, with June
and July having the highest reported number of cases across time. Multivariable analysis
revealed an age category of 5-17 and combined smoking/alcohol behaviors to be specific
risk factors for serotype 1 IPD. Serotype 1 IPD was a significant predictor of a clinical
diagnosis of meningitis (p<0.01) and LRTI (p=0.04) at discharge when controlling for
HIV infection at time of admission, underlying conditions other than HIV, antibiotic use,
age, race, and province.

Table of Contents

Table of Contents

I. Project Overview:10

II. Introduction:12

1. Justification for study:12

2. Literature review:12

2.1 Serotype-Specific Epidemiology - Invasiveness12

2.2 Antibiotic Resistance. 13

2.3 Clinical Syndrome. 14

2.4 Vaccines. 15

2.5 Age Specific Factors. 17

2.6 HIV.. 18

2.7 Outbreaks. 19

2.8 Anti-Retroviral Therapy and Cotrimoxazole prophylaxis20

2.9 Relationship Between Serotype and Clinical Outcomes, including Mortality and Severity of Disease 22

3. Intended/Potential Use of Study Findings:23

4. Study Design/Location:23

5. Objectives:24

6. Hypotheses:25

III. Procedures/Methods:28

1. Study Population:28

1.1 Description and source of study population28

1.2 Case Definition.. 29

1.3 Inclusion/exclusion criteria29

1.4 Sampling, including sample size and statistical power29

1.5 Consent process. 30

1.6 Audience and stakeholder participation30

2. Variables:31

2.1 Surveillance and Laboratory Methods34

3. Data Handling and Analysis:35

3.1 Analysis. 35

3.2 Quality Control/Assurance41

3.3 Faculty and Ethical Approval41

3.4 Handling of Adverse Events:41

IV Results:43

1. Descriptive Analysis. 43

1.1 Basic epidemiological characteristics43

1.2 Seasonality. 43

2. Bivariate Analysis. 44

2.1 Hypothesized unique risk factors for serotype 1 IPD44

2.2 Antiretroviral therapy and cotrimoxazole prophylaxis44

2.3 Clinical syndrome, disease severity, and mortality45

3. Multivariable analysis. 45

3.1 Initial Model for Serotype 1 IPD Risk Factors45

3.2 Confounding Assessment and Final Model46

3.3 Analysis of antiretroviral use among HIV positive patients with IPD infection.. 46

3.4 Multivariable models containing serotype 1 IPD as a predictor of clinical outcomes. 47

V Discussion:49

1. Strengths and Weaknesses49

2. Discussion of Findings. 50

VI Conclusion:56

VII References:57

VIII. Figures and Tables.. 65

Figure 1.. 65

Figure 2.. 65

Figure 3.. 66

Figure 4.. 66

Figure 5.. 67

Figure 6.. 67

Figure 7.. 68

Figure 8.. 72

Table 1.. 75

Table 2.. 76

Table 3.. 76

Table 4.. 77

Table 5.. 77

Table 6.. 78

Table 7.. 78

Table 8.. 79

Table 9.. 79

Table 10.. 80

About this Master's Thesis

Rights statement
  • Permission granted by the author to include this thesis or dissertation in this repository. All rights reserved by the author. Please contact the author for information regarding the reproduction and use of this thesis or dissertation.
  • English
Research Field
Committee Chair / Thesis Advisor
Committee Members
Partnering Agencies
Last modified

Primary PDF

Supplemental Files