Invasive Streptococcus pneumoniae infection among hospitalized patients in Jingzhou city, China, 2010-2012

Background Streptococcus pneumoniae (Sp) is a leading cause of bacterial pneumonia, meningitis, and sepsis and a major source of morbidity and mortality worldwide. Invasive pneumococcal disease (IPD) is defined as isolation of Sp from a normally sterile site, including blood or cerebrospinal fluid. The aim of this study is to describe outcomes as well as clinical and epidemiological characteristics of hospitalized IPD case patients in central China. Methods We conducted surveillance for IPD among children and adults from April 5, 2010 to September 30, 2012, in four major hospitals in Jingzhou City, Hubei Province. We collected demographic, clinical, and outcome data for all enrolled hospitalized patients with severe acute respiratory infection (SARI) or meningitis, and collected blood, urine, and cerebrospinal fluid (CSF) for laboratory testing for Sp infections. Collected data were entered into Epidata software and imported into SPSS for analysis. Results We enrolled 22,375 patients, including 22,202 (99%) with SARI and 173 (1%) with meningitis. One hundred and eighteen (118, 3%) with either SARI or meningitis were Sp positive, 32 (0.8%) from blood/CSF culture, and 87 (5%) from urine antigen testing. Of those 118 patients, 57% were aged ≥65 years and nearly 100% received antibiotics during hospitalization. None were previously vaccinated with 7-valent pneumococcal conjugate vaccine (PCV 7), 23-valent pneumococcal polysaccharide vaccine, or seasonal influenza vaccine. The main serotypes identified were 14, 12, 3, 1, 19F, 4, 5, 9V, 15 and 18C, corresponding to serotype coverage rates of 42%, 63%, and 77% for PCV7, PCV10, and PCV13, respectively. Conclusions Further work is needed to expand access to pneumococcal vaccination in China, both among children and potentially among the elderly, and inappropriate use of antibiotics is a widespread and serious problem in China.


Methods
We conducted surveillance for IPD among children and adults from April 5, 2010 to September 30, 2012, in four major hospitals in Jingzhou City, Hubei Province. We collected demographic, clinical, and outcome data for all enrolled hospitalized patients with severe acute respiratory infection (SARI) or meningitis, and collected blood, urine, and cerebrospinal fluid (CSF) for laboratory testing for Sp infections. Collected data were entered into Epidata software and imported into SPSS for analysis. PLOS

Introduction
Streptococcus pneumoniae (Sp) is a leading cause of bacterial pneumonia, meningitis, and sepsis and a major source of morbidity and mortality worldwide [1]. Although pneumococcal vaccines help prevent pneumococcal disease, the WHO estimates that approximately 1 million children die of pneumococcal disease every year, mostly in developing countries [2,3]. Although global estimates are valuable, national public health officials need country and region-specific estimates of disease incidence and prevalence for policy decision making [4].
To date, China has lacked substantial, high quality surveillance systems to measure incidence of Sp infection associated with respiratory disease and meningitis. Bacterial culture from blood or cerebrospinal fluid is the established method of diagnosing invasive Sp infection [5], but bacterial culture is infrequently used in China hospitals. Additionally, patients in China frequently take antibiotics by self-purchased before seeking medical care [6], potentially resulting in false negatives for Sp culture. The above factors make it challenging to estimate the burden of Sp infections in China.
In April 2010, the Chinese Center for Disease Control and Prevention (China CDC), in collaboration with the United States Centers for Disease Control and Prevention (US CDC), launched active surveillance for severe acute respiratory infection (SARI) and meningitis in Jingzhou city, Hubei province. The aim of this project was to characterize etiologies of SARI and meningitis in people admitted with these conditions to Jingzhou City hospitals. We analyzed data from this surveillance system to describe the epidemiology and clinical characteristics of SARI and meningitis patients with Sp infections.

Methods
We conducted surveillance in three general hospitals and one pediatric hospital in Jingzhou City, Hubei Province located in central China as previously described [7]. We collected demographic, clinical, and outcome data for all enrolled hospitalized patients with severe acute respiratory infection (SARI) or meningitis. Additionally, we obtained blood, urine, and cerebrospinal fluid (CSF) samples to test for Sp infections between April 5, 2010 and September 30, 2012.

Case definitions
Patients met the case definition for SARI if they were hospitalized and had: temperature !37.3˚C and at least one of the following: cough, sore throat, tachypnea, difficulty breathing, abnormal breath sounds on auscultation, sputum production, hemoptysis, chest pain, or chest radiograph consistent with pneumonia. Patients met the case definition for meningitis if they were hospitalized and had: temperature !37.3˚C and change in mental status, meningeal irritation signs (positive Kernig's or Brudzinski's signs), bulging fontanelle (if aged <12 months), toxic appearance, petechial rash, or purpural rash. Patients were excluded from the study if they were infants who were born in the hospital and became ill while hospitalized. A case was defined as laboratory-confirmed for Sp if it had a blood, urine (!18 years), or CSF specimen that was positive for Sp by bacterial culture or antigen testing.

Data collection
Physicians screened potential cases for SARI or meningitis at hospital admission, after obtaining verbal consent, or proxy consent (for children), physicians reviewed patients' medical records to obtain demographic information, past medical history including records on vaccinations, antibiotics use before hospitalization, clinical signs and symptoms, and radiographic results on a structured case report form. After discharge, physicians were also requested to update the case report form to complete data about treatment, disease outcomes and laboratory testing results. Patients were followed up to collect outcomes at 30 days after discharge.

Specimen collection and testing
For SARI patients, nurses collected whole blood specimens and urine specimens (!18 years) within two hours of enrollment or before the use of antibiotics following standardized procedures. Whole blood specimens (10ml for patients aged >15 years; 4 ml for those aged 7-14 years; 3 ml for those aged 3-7 years; 1 ml for those aged <3 years) were collected and then transferred to a hospital clinical laboratory within two hours to culture by BD-9050 automated culture system for Sp. For all SARI patients aged >18 years, nurses collected 10 ml of urine to detect Sp antigen. Urine specimens were tested by the Binax NOW Streptococcus pneumoniae urinary antigen test per the manufacturer's instructions. For meningitis patients, the collection and testing of blood samples was the same as for SARI patients. Cerebrospinal fluid (CSF) was obtained within two hours of enrollment or before initiation of antibiotic treatment following standard procedures for lumbar puncture; 40 µl of CSF was inoculated into bacterial culture within one hour of collection. For Sp isolates, PCR was used to determine serotype using previously published methods [8].

Data analysis
Data were entered into Epidata software (version 3.1) and imported into SPSS (v17.0, SPSS, Chicago, IL, USA) for analysis. To describe baseline characteristics, clinical, treatment, complication, and outcome, we conducted frequency analyses for categorical variables and estimated median and interquartile ranges (IQRs) for continuous variables. Patients with laboratoryconfirmed Sp were divided into age groups (<5 years, 5-64 years group, and aged !65 years) and compared using chi-squared test and Fisher's exact test to compare categorical variables, and t-test to compare continuous variables; all statistical tests were two-sided with p<0.05.

Human subjects review
This project was approved by the ethical review committees at the Chinese Center for Disease Control and Prevention (China CDC, Beijing, China) and the U.S. Centers for Disease Control and Prevention (US CDC, Atlanta, GA, USA). In response to the 2009 H1N1 pandemic, in October 2009 China's Ministry of Health issused the national surveillance protocal for SARI and authorized participating hospitals to collect individual case data and specimens from SARI patients. Therefore, for SARI patients in this study, participation only required patients or their parent/guardian to provide brief verbal consent. Blood culture and lumbar puncuture for suspected meningitis are the routine clinical practice in hospitals, consent in China for such procedures is almost always verbal, not written.

Results
A total of 22,608 hospitalized patients in the four surveillance hospitals met either the SARI or meningitis case definitions from April 5, 2010 to September 30, 2012. Two hundred and thirty three (n = 233) refused to participate. We enrolled 22,375 total patients: 22,202 (99%) SARI and 173 (1%) meningitis.

Discussion
In this large, prospective study in China, Sp was infrequently identified as the cause of pneumonia and meningitis, and few patients were vaccinated against Sp infection.
Multiple studies have been published in China documenting Sp as a common cause of respiratory infection in children and adults, but most relied on the use of non-sterile specimens, such as oro-pharyngeal or naso-pharyngeal swabs [5,[9][10][11][12]. Our study, in contrast, is the first large study in China to prospectively enroll patients with a broad case definition of severe acute respiratory infection or meningitis and to test specimens from sterile sites. Sp is considered the most common cause of bacterial community acquired pneumonia (CAP) [13], and may cause 30-50% of pneumonia requiring hospitalization in Europe and United States [14]. Although 72% of patients with Sp detected in our study were diagnosed with pneumonia, only 7% of patients diagnosed with pneumonia had Sp detected. Sp is also a common cause of meningitis globally [15], and was the most common cause of bacterial meningitis in one previous population-based study in China [16]. Nevertheless, Sp was only identified in 6% of patients who met our case definition of meningitis.
The low rate of Sp detection in our study may be due to the expansive case definition. First, we intentionally enrolled a broad cross-section of patients with severe illness, which resulted in many patients in whom Sp infection was unlikely to be detected, such as adults aged 18-64 years. Second, over 50% of patients received antibiotics before hospitalization, increasing the  (Continued) likelihood that bacterial cultures of blood and CSF would be negative [17][18][19][20]. Third, we embedded the study within routine clinical practice, which could have limited the quality of specimen collection and testing. The optimum time for specimen collection is actually before symptoms occur [21], but in our study, specimens were not collected until hospitalization, which occurred a median of 2 days after symptom onset. Lastly, the methods to detect pneumococcus, such as real-time PCR or Binax NOW1 antigen test, can dramatically increase the detection of pneumococcus, especially in thoese already having received antibiotics. Although real-time PCR is a more sensitive method of detecting invasive pneumococcal disease, bacterial culture is the standard diagnostic method that has been used in local hospitals for years and there were insufficient funds to support this method. WHO and ACIP recommend the use of conjugate pneumooccal vaccine to prevent invasive disease in children and the elderly [2,22]. We found that no patients with Sp detected reported previous Sp vaccination. Sp vaccination is uncommon in China, because it is not included in the national immunization program for children or adults, to the relatively low awareness of the disease, and to the high out-of-pocket charges for Sp vaccine [23]. In our study, the median age of patients with Sp was 68 years, consistent with studies indicating the elderly are at high risk for invasive pneumococcal disease and can benefit from vaccination [24].
Despite their age and the frequency of comorbid medical conditions, few patients with Sp died in our study. There are other reasons such as prompt and supportive medical care, but one possibility is that early, widespread use of antibiotics prevented death [25]. We found that PCV13, which has recently been licensed in China, would have covered 77% of the Sp infections identified detected in our study, including 68% of those in people aged <5 years and 88% in those aged !5 years. WHO recommends that pneumococcal conjugate vaccine be included in national childhood immunization programs, but does not make a similar recommendation for the elderly [2]. Given the population-level and individual harms associated with widespread antibiotic use, the documented benefits of pneumococcal vaccination in other settings [1,21,26,27], and the expected serotype coverage of PCV13 in China, further work is needed to expand access to pneumococcal vaccination in China, both among children and potentially among the elderly.
In China, inappropriate use of antibiotics is a widespread and serious problem [28,29]. In our study, over 50% of enrolled patients received antibiotic treatment before hospitalization, and nearly 100% of enrolled patients received antibiotics during hospitalization, even though few had a documented bacterial infection. Widespread use of antibiotics may increase resistance and lead to individual health complications. China has recently begun programs to promote judicious prescribing of antibiotics and to restrict pharmacies from dispensing antibiotics without a doctor's prescription. Additionally, corticosteroid treatment has been used widely as adjunctive treatment for adults with bacterial meningitis [30]. About 80% of meningitis patients with confirmed Sp infection received corticosteroid treatment. Corticosteroid treatment has  been independently associated with favorable outcomes and survival in patients with pneumococcal and non-pneumococcal meningitis [31,32]. In addition to the limitations noted above, our study was also limited to only one city, did not enroll outpatients, and relied on routinely collected data, albeit prospectively, from medical records. It is possible that we, therefore, under-estimated incidence and disease burden. Nevertheless, our study adds to the limited literature about Sp infections in China and provides further data to support the potential benefits of Sp vaccination among children and the elderly Invasive pneumococcal infections based on surveillance in China. Furthermore, the characterization of Sp isolates and infections in this study can contribute to more rapid detection and control of Sp outbreaks in China, and thereby contribute to enhanced global health security.
Supporting information S1