Publications by authors named "Kendra M Viner"

13 Publications

  • Page 1 of 1

Association between discharges against medical advice and readmission in patients treated for drug injection-related skin and soft tissue infections.

J Subst Abuse Treat 2021 07 7;126:108465. Epub 2021 May 7.

Division of Substance Use Prevention and Harm Reduction, Philadelphia Department of Public Health, 123 South Broad Street, Philadelphia, PA 19109, USA. Electronic address:

Background: The prevalence of injection drug use (IDU)-related skin and soft tissue infections (SSTI) in Philadelphia has been steadily increasing since 2013. Patients seeking treatment for these infections are more likely to be discharged against medical advice (AMA), increasing the likelihood that they will end antibiotic treatment prematurely and require additional medical interventions.

Methods: The research team performed a nested case-control study using the Pennsylvania Health Care Cost Containment Council database for Philadelphia residents hospitalized for SSTI and substance use-related diagnoses between 2013 and 2018. The primary outcome was readmission in the same or following quarter. The study examined the impact of discharge AMA on readmission along with clinical characteristics including diagnoses for anxiety, bipolar disorder, depression, schizophrenia, diabetes, and polydrug use.

Results: There were 8265 hospitalizations for IDU-related SSTI and 316 (6%) were readmitted to the hospital at least once in the same or following quarter. In total, 23.4% of cases and 13% of controls left AMA. In the final multivariable regression model, AMA discharge (AOR 2.04, 95% CI 1.46-2.86), anxiety (AOR 1.44, 95% CI 1.01-2.05), diabetes (AOR 2.02, 95% CI 1.46-2.81), and polydrug use (AOR 2.11, 95% CI 1.52-2.92) were associated with higher odds of readmission.

Conclusions: Our study demonstrates that readmissions for IDU-related SSTI are associated with recent discharge AMA. As IDU-related SSTI and polydrug use continue to rise, premature antibiotic treatment completion will impact more people, leading to worse health outcomes and additional strain on the health care system.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1016/j.jsat.2021.108465DOI Listing
July 2021

Adherence to buprenorphine: An analysis of prescription drug monitoring program data.

Drug Alcohol Depend 2020 11 28;216:108317. Epub 2020 Sep 28.

Philadelphia Department of Public Health, Division of Substance Use Prevention and Harm Reduction, 123 S. Broad St., Suite 1120, Philadelphia, PA 19109, USA. Electronic address:

Background: Although buprenorphine is an evidence-based treatment for opioid use disorder (OUD), many individuals discontinue treatment soon after starting. This study assesses predictors of buprenorphine adherence using Prescription Drug Monitoring Program (PDMP) data.

Methods: PDMP data for Philadelphia, Pennsylvania were used to measure 180-day adherence to buprenorphine among new initiates. Adherence was classified using percent days covered (PDC), and new initiates with PDC ≥ 0.80 were classified as adherent. Multivariable logistic regression was conducted to determine factors associated with buprenorphine adherence.

Results: Between January 2017 and December 2018, 10,669 Philadelphia residents initiated buprenorphine and 26.6 % remained adherent after 180 days. Demographic factors associated with greater odds of adherence included age category and female sex (aOR: 1.37; 95 % CI: 1.25-1.50). Those filling an opioid prescription, other than buprenorphine, during the follow-up period had lower odds of adherence than those who did not fill an opioid prescription (aOR: 0.62; 95 % CI: 0.50-0.77). Odds of adherence was greater for those on the film formulation (aOR: 1.37; 95 % CI: 1.25-1.50) than the tablet formulation. Those filling medium (aOR: 1.76; 95 % CI: 1.55-2.00) and high dose (aOR: 5.11; 95 % CI: 4.30-6.17) buprenorphine prescriptions had higher odds of adherence than those filling low dose prescriptions.

Conclusions: Individual demographics, receipt of an opioid prescription, buprenorphine formulation, and buprenorphine dose were all associated with adherence to buprenorphine. Ongoing strategies to address OUD need to prioritize increasing retention in long-term evidence-based buprenorphine treatment while also encouraging providers to regularly consult the PDMP to ensure patient compliance.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1016/j.drugalcdep.2020.108317DOI Listing
November 2020

Correlates of experiencing and witnessing non-fatal opioid overdoses among individuals accessing harm reduction services in Philadelphia, Pennsylvania.

Subst Abus 2020 23;41(3):301-306. Epub 2019 Oct 23.

Philadelphia Department of Public Health, Office of the Health Commissioner, Philadelphia, PA, USA.

: With the dramatic increase in overdose deaths in the United States, layperson overdose education and naloxone distribution (OEND) programs serve a critical role in preventing opioid overdose fatality. This study examines associations for witnessing an opioid overdose or experiencing a non-fatal opioid overdose to identify new opportunities for expansion of OEND programs. : Cross sectional surveys were administered at the sole needle and syringe exchange program in Philadelphia, PA. Bivariate and multivariable logistic regression analyses were conducted to examine associations for witnessing an opioid overdose or experiencing a non-fatal opioid overdose in the previous 12 months. : In total, 370 individuals were identified as using opioids in the previous three months and included in the study. Factors associated with experiencing a non-fatal opioid overdose were unstable housing (aOR: 2.16; 95% CI: 1.12-3.99), recent incarceration (aOR: 2.34; 95% CI: 1.41-3.89), co-use of opioids and benzodiazepines (aOR: 2.04; 95% CI: 1.22-3.41), co-use of heroin and cocaine (aOR: 1.69; 95% CI: 1.04-2.75), drug injection (aOR: 4.25; 95% CI: 1.90-9.54), inpatient detoxification history (aOR: 2.33; 95% CI:1.27-4.43), and witnessing an overdose in the previous 12 months (aOR: 2.51; 95% CI: 1.02-6.13). Factors associated with witnessing an overdose were unstable housing (aOR: 5.14; 95% CI: 2.57-10.28), recent incarceration (aOR: 2.88; 95% CI: 1.24-6.74), and a history of being trained to use naloxone (aOR: 3.39; 95% CI: 1.63-7.04). Findings presented here show characteristics of those most likely to witness an overdose or experience a non-fatal overdose who could be served by expansion of OEND programs.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1080/08897077.2019.1675115DOI Listing
September 2021

Beyond the walls: Risk factors for overdose mortality following release from the Philadelphia Department of Prisons.

Drug Alcohol Depend 2018 08 5;189:108-115. Epub 2018 Jun 5.

Philadelphia Department of Public Health, Opioid Surveillance, Epidemiology, and Prevention Program, 1101 Market St., Suite 1320, Philadelphia, PA 19107, USA.

Background: High overdose mortality after release from state prison systems is well documented; however, little is known about overdose mortality following release from local criminal justice systems (CJS). The purpose of this study was to assess overdose mortality following release from a local CJS in Philadelphia, PA.

Methods: We conducted a retrospective cohort study of individuals released to the community from a local CJS between 2010 and 2016. Incarceration records were linked to overdose fatality data from the Medical Examiner's Office and death certificate records. All-cause, overdose, and non-overdose mortality were examined.

Results: Of the 82,780 individuals released between 2010 and 2016, 2,522 (3%) died from any cause, of which 837 (33%) succumbed to overdose. Individuals released from incarceration had higher risk of overdose death compared to the non-incarcerated population (Standardized Mortality Ratio [SMR]: 5.29, 95% CI 4.93-5.65), and risk was greatest during the first two weeks following release (SMR: 36.91, 95% CI: 29.92-43.90). Among released individuals, black, non-Hispanic individuals (Hazard Rate [HR]: 0.17, 95% CI: 0.14-0.19) and Hispanic individuals (HR: 0.41, 95% CI: 0.34-0.50) were at lower risk for overdose than white, non-Hispanic individuals. Individuals released with a serious mental illness (SMI) were at higher risk of overdose (HR: 1.54, 95% CI: 1.27-1.87) than those without a SMI.

Discussion: Previously incarcerated individuals are at high risk of overdose death following release from a local CJS, especially in the earliest weeks following release. Prevention measures including behavioral health treatment and referral and take-home naloxone may reduce overdose mortality after release.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1016/j.drugalcdep.2018.04.034DOI Listing
August 2018

Influence of birth origin and risk factor profile on hepatitis B mortality: Philadelphia, PA 2003-2013.

Ann Epidemiol 2018 03 22;28(3):169-174. Epub 2017 Dec 22.

Division of Disease Control, Philadelphia Department of Public Health, Philadelphia, PA.

Purpose: Chronic hepatitis B virus (HBV) affects specific subpopulations in the United States, including individuals born in HBV-endemic countries and persons engaging in high-risk behaviors.

Methods: The 2003-2013 HBV registry data and surveillance investigations for Philadelphia, PA were matched to death certificate data to examine demographic, risk factor, and cause of death characteristics among HBV-infected populations. Bivariate analysis compared investigated foreign-born (FB) and US-born chronic HBV individuals. Multivariable logistic regression assessed associations between HBV-status, birth origin, demographic information, and liver-related death.

Results: Of 773 investigated HBV-infected individuals, 159 were US-born and 614 were FB and of primarily non-Hispanic Asian descent. Behavioral risk factors were more often reported by US-born individuals. HBV-infected FB decedents were twice as likely as US-born decedents to have a liver-related cause of death, whereas HIV/AIDS and drug overdose were more likely causes of death among those born in the United States.

Conclusions: There are two HBV-infected populations in Philadelphia: 1) FB individuals most likely infected at birth or during early childhood and 2) US-born individuals with behaviors suggestive of risk-related HBV acquisition. These findings illustrate the need for both FB and US-born individuals with ongoing risk behaviors to receive routine HBV screening, vaccination if indicated, and medical care for outcomes of chronic HBV infection.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1016/j.annepidem.2017.12.006DOI Listing
March 2018

Capture-Recapture: Using Existing Data Sources to Improve Perinatal Hepatitis B Surveillance, Philadelphia, 2008-2014.

Public Health Rep 2017 May/Jun;132(3):376-380. Epub 2017 Apr 13.

1 Division of Disease Control, Philadelphia Department of Public Health, Philadelphia, PA, USA.

Objective: The objective of this study was to describe the capture-recapture method used by the Philadelphia Department of Public Health to enhance surveillance of perinatal hepatitis B virus (HBV), report on results and limitations of the process, and determine why some HBV-positive mother-infant pairs were not initially identified by Philadelphia's Perinatal Hepatitis B Prevention Program (PHBPP).

Methods: We performed capture-recapture retrospectively for births in 2008 and 2009 in Philadelphia and prospectively for births from 2010 to 2014 by independently matching annual birth certificate data to PHBPP and HBV surveillance data. We compared the number of HBV-positive mother-infant pairs identified each year to the point estimates and lower-limit estimates calculated by the Centers for Disease Control and Prevention for the Philadelphia PHBPP.

Results: Of 156 605 pregnancy outcomes identified between 2008 and 2014, we found 1549 HBV-positive mother-infant pairs. Of 705 pairs that were initially determined, 358 (50.7%) were confirmed to be previously unidentified HBV-positive pairs. Reasons for failing to identify these mother-infant pairs prior to capture-recapture included internal administrative issues (n = 191, 53.4%), HBV testing and reporting issues (n = 92, 25.7%), and being lost to follow-up (n = 75, 20.9%). Each year that capture-recapture was used, the number of pairs identified by the Philadelphia PHBPP exceeded the Centers for Disease Control and Prevention's lower-limit estimates for HBV-positive mother-infant pairs.

Conclusions: Capture-recapture was useful for identifying HBV-positive pregnant women for Philadelphia's PHBPP and for highlighting inadequacies in health department protocols and HBV testing during pregnancy. Other health departments with access to similar data sources and staff members with the necessary technical skills can adapt this method.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1177/0033354917702851DOI Listing
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC5415262PMC
June 2017

Failure to Test and Identify Perinatally Infected Children Born to Hepatitis C Virus-Infected Women.

Clin Infect Dis 2016 Apr 20;62(8):980-5. Epub 2016 Jan 20.

Division of Disease Control, Philadelphia Department of Public Health, Pennsylvania.

Background: Vertical transmission of hepatitis C virus (HCV) is the most common route of pediatric HCV infection. Approximately 5% of children born to HCV-infected mothers develop chronic infection. Recommendations employ risk-based HCV testing of pregnant women, and screening children at a young age. This study assesses testing rates of children born to mothers tested HCV-positive in a major US city with a high burden of HCV infection.

Methods: HCV surveillance data reported to the Philadelphia Department of Public Health are housed in the Hepatitis Registry. Additional tests, including negative results, were retrospectively collected. HCV data were matched with 2011-2013 birth certificates of children aged ≥20 months to identify mothers tested HCV-positive and screened children. The observed perinatal HCV seropositivity rate was compared to the expected rate (5%).

Results: A total of 8119 females aged 12-54 years tested HCV-positive and in the Hepatitis Registry. Of these, 500 (5%) had delivered ≥1 child, accounting for 537 (1%) of the 55 623 children born in Philadelphia during the study period. Eighty-four (16%) of these children had HCV testing; 4 (1% of the total) were confirmed cases. Twenty-three additional children are expected to have chronic HCV infection, but were not identified by 20 months of age.

Conclusions: These findings illustrate that a significant number of women giving birth in Philadelphia test positive for HCV and that most of their at-risk children remain untested. To successfully identify all HCV-infected children and integrate them into HCV-specific care, practices for HCV screening of pregnant women and their children should be improved.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1093/cid/ciw026DOI Listing
April 2016

Comparison of risk-based hepatitis C screening and the true seroprevalence in an urban prison system.

J Urban Health 2015 Apr;92(2):379-86

Philadelphia Department of Public Health, Division of Disease Control, 500 S. Broad St., Philadelphia, PA, 19146, USA,

Hepatitis C virus (HCV) is the most common blood-borne infection in the USA, though seroprevalence is elevated in certain high-risk groups such as inmates. Correctional facility screening protocols vary from universal testing to opt-in risk-based testing. This project assessed the success of a risk-based HCV screening strategy in the Philadelphia Prison System (PPS) by comparing results from current testing practices during 2011-2012 (Risk-Based Screening Group) to a September 2012 blinded seroprevalence study (Philadelphia Department of Public Health (PDPH) Study Cohort). PPS processed 51,562 inmates in 2011-2012; 2,727 were identified as high-risk and screened for HCV, of whom 57 % tested HCV antibody positive. Twelve percent (n = 154) of the 1,289 inmates in the PDPH Study Cohort were anti-HCV positive. Inmates ≥30 years of age had higher rates of seropositivity in both groups. Since only 5.3 % of the prison population was included in the Risk-Based Screening Group, an additional 4,877 HCV-positive inmates are projected to have not been identified in 2011-2012. Gaps in case identification exist when risk-based testing is utilized by PPS. A more comprehensive screening model such as opt-out universal testing should be considered to identify HCV-positive inmates. Identification of these individuals is an important opportunity to aid underserved high-risk populations and to provide medical care and secondary prevention.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1007/s11524-015-9945-4DOI Listing
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC4411325PMC
April 2015

Knowledge, attitudes, and practices for diagnosing breakthrough varicella in the outpatient setting.

Public Health Rep 2012 Nov-Dec;127(6):585-90

Philadelphia Department of Public Health, Division of Disease Control, Philadelphia, PA, USA.

Objectives: We assessed provider knowledge, attitudes, and practices for the management of breakthrough varicella and identified barriers to implementation of laboratory testing and reporting.

Methods: We surveyed 145 health-care providers (HCPs) from 30 pediatric practices in Philadelphia who did not have a history of laboratory testing for breakthrough varicella. The self-administered survey instrument collected information on clinicians' practices for management of children presenting with rash, infection-control strategies, reporting to public health agencies, and laboratory testing.

Results: Among the 144 HCPs who completed the survey, 73 (51%) had practiced for more than 10 years. While 115 HCPs (80%) would elect to evaluate a child with rash in the office, only 19 (13%) would submit diagnostics. When patients had a known recent exposure to varicella, 84 HCPs (58%) would use laboratory tests: 40% would use direct fluorescent antibody staining on a specimen from a cutaneous lesion, 24% would use polymerase chain reaction on a lesion specimen, 21% would use acute and convalescent serology, and 10% would use other tests. While waiting for test results, 82 HCPs (57%) would advise that the child be kept at home, 39 (27%) would notify the local health department, and 33 (23%) would inform the school nurse.

Conclusion: As varicella becomes increasingly uncommon, laboratory confirmation becomes more critical for appropriate diagnosis, similar to poliomyelitis and measles. Our findings suggest that HCPs need further education regarding laboratory confirmation, containment, and reporting of breakthrough varicella.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1177/003335491212700608DOI Listing
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3461849PMC
January 2013

The Vaccinia virus complement control protein modulates adaptive immune responses during infection.

J Virol 2011 Mar 29;85(6):2547-56. Epub 2010 Dec 29.

University of Pennsylvania School of Medicine, Division of Infectious Diseases, 502 Johnson Pavilion, Philadelphia, PA 19104-6073, USA.

Complement activation is an important component of the innate immune response against viral infection and also shapes adaptive immune responses. Despite compelling evidence that complement activation enhances T cell and antibody (Ab) responses during viral infection, it is unknown whether inhibition of complement by pathogens alters these responses. Vaccinia virus (VACV) modulates complement activation by encoding a complement regulatory protein called the vaccinia virus complement control protein (VCP). Although VCP has been described as a virulence factor, the mechanisms by which VCP enhances VACV pathogenesis have not been fully defined. Since complement is necessary for optimal adaptive immune responses to several viruses, we hypothesized that VCP contributes to pathogenesis by modulating anti-VACV T cell and Ab responses. In this study, we used an intradermal model of VACV infection to compare pathogenesis of wild-type virus (vv-VCPwt) and a virus lacking VCP (vv-VCPko). vv-VCPko formed smaller lesions in wild-type mice but not in complement-deficient mice. Attenuation of vv-VCPko correlated with increased accumulation of T cells at the site of infection, enhanced neutralizing antibody responses, and reduced viral titers. Importantly, depleting CD8(+) T cells together with CD4(+) T cells, which also eliminated T helper cell-dependent Ab responses, restored vv-VCPko to wild-type levels of virulence. These results suggest that VCP contributes to virulence by dampening both antibody and T cell responses. This work provides insight into how modulation of complement by poxviruses contributes to virulence and demonstrates that a pathogen-encoded complement regulatory protein can modulate adaptive immunity.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1128/JVI.01474-10DOI Listing
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3067952PMC
March 2011

Cell surface expression of the vaccinia virus complement control protein is mediated by interaction with the viral A56 protein and protects infected cells from complement attack.

J Virol 2008 May 20;82(9):4205-14. Epub 2008 Feb 20.

University of Pennsylvania School of Medicine, Division of Infectious Diseases, 502 Johnson Pavilion, Philadelphia, PA 19104-6073, USA.

The vaccinia virus (VACV) complement control protein (VCP) is the major protein secreted from VACV-infected cells. It has been reported that VCP binds to the surfaces of uninfected cells by interacting with heparan sulfate proteoglycans (HSPGs). In this study, we show that VCP is also expressed on the surfaces of infected cells and demonstrate that surface localization occurs independently of HSPGs. Since VCP does not contain a transmembrane domain, we hypothesized that VCP interacts with a membrane protein that localizes to the infected-cell surface. We show that the VACV A56 membrane protein is necessary for the cell surface expression of VCP and demonstrate that VCP and A56 interact in VACV-infected cells. Since the surface expression of VCP was abrogated by reducing agents, we examined the contribution of an unpaired cysteine residue on VCP to VCP surface expression and VCP's interaction with A56. To do this, we mutated the unpaired cysteine in VCP and generated a recombinant virus expressing the altered form of VCP. Following the infection of cells with the mutant virus, VCP was neither expressed on the cell surface nor able to interact with A56. Importantly, the cell surface expression of VCP was found to protect infected cells from complement-mediated lysis. Our findings suggest a new function for VCP that may be important for poxvirus pathogenesis and impact immune responses to VACV-based vaccines.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1128/JVI.02426-07DOI Listing
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2293032PMC
May 2008

B5-deficient vaccinia virus as a vaccine vector for the expression of a foreign antigen in vaccinia immune animals.

Virology 2007 May 26;361(2):356-63. Epub 2006 Dec 26.

Department of Medicine, University of Pennsylvania School of Medicine, Division of Infectious Diseases, Philadelphia, PA 19104-6073, USA.

Recombinant vaccinia viruses have shown promise as vaccine vectors. However, their effectiveness is markedly reduced by pre-existing anti-vaccinia immunity. The possibility of new vaccinia immunizations in the event of a bioterror-related smallpox release poses an additional negative impact on the utility of vaccinia-based vectors. Thus, we aimed to design a vaccinia vector that would enhance the immune response to an expressed foreign protein in a pre-immune animal model. To do this, we made use of the finding that most neutralizing antibodies against the extracellular form of vaccinia virus are directed against the B5 protein. We found that mice immunized with vaccinia, primed with Gag plasmid DNA, and boosted with a recombinant vaccinia virus lacking the majority of the B5 ectodomain expressing a test antigen, HIV Gag, had stronger anti-Gag immune responses than mice that were boosted with a wild-type virus-expressing Gag. These findings are particularly striking given the more attenuated phenotype of this virus, as compared to its wild-type counterpart. Importantly, we found that vaccination with a B5R deletion virus, followed by boosting with the Gag-expressing virus lacking the majority of the B5 ectodomain, resulted in poorer anti-Gag immune responses. Thus, recombinant vaccinia viruses lacking the B5 ectodomain may serve as vaccine vectors in DNA prime-vaccinia boost vaccinations of individuals with pre-existing immunity against vaccinia. These data open the possibility of extending the potential benefit of replication competent recombinant vaccinia virus vectors to a larger population.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1016/j.virol.2006.11.020DOI Listing
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2048764PMC
May 2007

Activity of vaccinia virus-neutralizing antibody in the sera of smallpox vaccinees.

Microbes Infect 2005 Apr 22;7(4):579-83. Epub 2005 Mar 22.

Department of Medicine, Division of Infectious Diseases, University of Pennsylvania School of Medicine, 502 Johnson Pavilion, Philadelphia, PA 19104-6073, USA.

Individuals vaccinated against smallpox maintain substantial antiviral antibody responses for many years after vaccination. In this study, we examined the ability of antiviral antibodies from 104 unique serum samples to neutralize the two infectious forms of vaccinia virus, intracellular mature virus (IMV) and extracellular enveloped virus (EEV). While we found direct correlations between antiviral antibody titers and the ability to neutralize IMV and EEV, correlation with EEV neutralization was weaker. To determine factors that may influence more varied EEV neutralization within a vaccinated population, we asked the following questions. (1) Does vaccinia virus-neutralizing ability remain constant over time? (2) Do multiple vaccinations boost IMV and EEV neutralization activity? We found that serum from vaccinated individuals retained ability to neutralize EEV for a relatively long time, but there was a significant drop in EEV neutralization ability in the third decade after vaccination. While all vaccinees maintained some ability to neutralize IMV, a number of individuals lost the capacity to neutralize EEV. Interestingly, the ability to neutralize either virus form was not altered by the number of vaccinations received. Since it is likely that neutralizing antibodies against both IMV and EEV are required for maximal protective immunity, a loss of anti-EEV-neutralizing ability may warrant the revaccination of individuals who have been vaccinated >20 years ago, should widespread pre-event smallpox vaccination be instituted.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1016/j.micinf.2005.02.004DOI Listing
April 2005
-->