Emergent and re-emergent challenges in the theory of infectious diseases South Africa June, 2007 www.noveltp.com 1 The theory of infectious diseases has a rich history Sir Ronald Ross 1857-1932 2 But prediction is difficult • Disease systems are complex, characterized by nonlinearities and sudden flips image.guardian.co.uk/ 3 • They also are complex adaptive systems, integrating phenomena at multiple scales a dna ™emiTkciuQ rosserpmoced )desserpmocnU( F FIT .erutcip siht ees ot dedeen era QuickTime™ and a TIFF (Uncompressed) decompressor are needed to see this picture. lshtm.ac.uk encarta.msn.com Qu ic kTime ™ a nd a TIFF (Unc omp res se d) d ec omp res so r a re n ee de d to se e t his p ic tu re. www.who.int www.nobel.org 4 Integrating these multiple scales is one major challenge • • • • • • Pathogen Host individual Host population dynamics Pathogen genetics Host genetics Vector 5 Despite a century of elegant theory, new diseases emerge, old reemerge 6 http://edie.cprost.sfu.ca/gcnet Significant management puzzles remain • Whom should we vaccinate? QuickTime™ and a TIFF (Uncompressed) decompressor are needed to see this picture. 7 www.calcsea.org Whom should we vaccinate? • Those at greatest risk? QuickTime™ and a TIFF (Uncompressed) decompressor are needed to see this picture. QuickTime™ and a TIFF (Uncompressed) decompressor are needed to see this picture. www.nursingworld.org 8 Whom should we vaccinate? • Or those who pose greatest risk to others? QuickTime™ and a TIFF (Uncompressed) decompressor are needed to see this picture. 9 www-personal.umich.edu/~mejn Other management puzzles: Problems of the Commons • Individual benefits/costs vs. group benefits/costs – Vaccination – Antibiotic use • Hospitals and nursing homes vs. health-care providers vs. individuals These introduce game-theoretic problems 10 Antibiotic resistance threatens the effectiveness of our most potent weapons against bacterial infections QuickTime™ and a TIFF (Uncompressed) decompressor are needed to see this picture. 11 Lecture outline • Periodicities and fluctuations • Antibiotic resistance and other problems of the Commons 12 Many important diseases exhibit oscillations on multiple temporal and spatial scales 13 Measles in the U.K.; Grenfell et al. 2001 (Nature) QuickTime™ and a BMP decompressor are needed to see this picture. 14 Control must deal with temporal and spatial fluctuations QuickTime™ and a Cinepak decompressor are needed to see this picture. 15 Influenza global spread 16 Influenza A reemerges year after year, despite the fact that infection leads to lifetime immunity to a strain 17 18 U.S. mortality in the 20th century 19 The “Spanish Flu” of 1918 20 21 22 Bush, Fitch, Cox Timeseries of viral clusters 23 Fluctuations in influenza A • Rapid replacement at level of individual strains • Gradual replacement at level of subtypes • Recurrence at level of clusters 24 Standard SIR Model (No latency) Susceptible S Infectious I Removed R 25 Simplest model dI /dt SI I I deaths recovered 26 For spread: S Rs 1 Condition for spread in a naïve population 1 R 0 N 1 Secondary Average infections/ infectious time period Thus R0 is the #secondary/primary infection. 27 Interpretation if threshold is exceeded 1. With no new recruits, outbreak and collapse 2. With new births, get stable equilibrium 3. Oscillations require a more complicated model 28 Complications QuickTime™ and a TIFF (Uncompressed) decompressor are needed to see this picture. • New immigrants H.M.S. Bounty www.lareau.org 29 Complications QuickTime™ and a TIFF (Uncompressed) decompressor are needed to see this picture. • New immigrants • Demography www.lareau.org 30 Complications • New immigrants • Demography • Heterogeneous mixing patterns QuickTime™ and a TIFF (Uncompressed) decompressor are needed to see this picture. www.lareau.org 31 Complications • • • • New immigrants Demography Heterogeneous mixing patterns Genetic changes in host QuickTime™ and a TIFF (Uncompressed) decompressor are needed to see this picture. www.lareau.org 32 Complications • • • • • New immigrants Demography Heterogeneous mixing patterns Genetic changes in host Multiple strains/diseases QuickTime™ and a TIFF (Uncompressed) decompressor are needed to see this picture. www.lareau.org 33 Complications • • • • • • New immigrants Demography Heterogeneous mixing patterns Genetic changes in host Multiple strains/diseases Vectors QuickTime™ and a TIFF (Uncompressed) decompressor are needed to see this picture. www.lareau.org 34 Complications • • • • • • New immigrants Demography Heterogeneous mixing patterns Genetic changes in host Multiple strains/diseases Vectors QuickTime™ and a TIFF (Uncompressed) decompressor are needed to see this picture. www.lareau.org 35 Oscillations • Stochastic factors • Seasonal forcing (e.g., in transmission rates) • Long periods of temporary immunity • Other explicit delays (e.g., incubation periods) • Age structure • Non-constant population size • Non-bilinear transmission coefficients • Interactions with other diseases/strains 36 Oscillations • Stochastic factors • Seasonal forcing (e.g., in transmission rates) • Long periods of temporary immunity • Other explicit delays (e.g., incubation periods) • Age structure • Non-constant population size • Non-(bilinear) transmission coefficients • Interactions with other diseases/strains 37 Oscillations • Stochastic factors • Seasonal forcing (e.g., in transmission rates) • Long periods of temporary immunity • Other explicit delays (e.g., incubation periods) • Age structure • Non-constant population size • Non-(bilinear) transmission coefficients • Interactions with other diseases/strains 38 Oscillations • Stochastic factors • Seasonal forcing (e.g., in transmission rates) • Long periods of temporary immunity • Other explicit delays (e.g., incubation periods) • Age structure • Non-constant population size • Non-(bilinear) transmission coefficients • Interactions with other diseases/strains 39 Oscillations • Seasonal forcing (e.g., in transmission rates) – Can interact with endogenous oscillations to produce chaos • Age structure – Creates implicit delays • Interactions with other diseases/strains – Includes, therefore, genetic change in pathogen 40 Interacting strains or diseases Susceptible Infectious 1 Recovered 1 Infectious 2 Infectious 2 R1 Recovered 2 Infectious 1 R2 Recovered 1,2 41 Understanding endogenous oscillations • Age-structured models can produce damped oscillations (Schenzele, Castillo-Chavez et al.) • Two-strain models can produce damped oscillations (Castillo-Chavez et al.) • Coupling these may lead to sustained periodic or other oscillations 42 Summary: Understanding endogenous oscillations • Age-structure • Epidemiology • Genetics all have characteristic scales of oscillation that can interact with each other, and with seasonal forcing 43 Lecture outline • Periodicities and fluctuations • Antibiotic resistance and other problems of the Commons 44 Problems of The Commons • Fisheries • Aquifers • Pollution QuickTime™ and a TIFF (Uncompressed) decompressor are needed to see this picture. 45 www.aisobservers.com Problems of The Commons • • • • Fisheries Aquifers Pollution Vaccines Quick Time™ and a TIFF (Uncompressed) dec ompressor are needed to s ee this pic ture. pubs.acs.org 46 images.usatoday.com Problems of The Commons • • • • • Fisheries Aquifers Pollution Vaccines Antibiotics 47 www.bath.ac.uk Antibiotic resistance is on the rise 48 www.wellcome.ac.uk 49 Would you deny your child antibiotics to maintain global effectiveness? 50 Antibiotic resistance is an increasing problem We are rapidly losing the benefits antibiotics have given us against a wide spectrum of diseases 51 52 53 Reasons for rise of antibiotic resistance • Agricultural uses • Overuse by physicians • Hospital spread (nosocomial infections) www.history.navy.mil/ac 54 Huang et al, Emerging Infectious Diseases, 2002 Hospitals are a major source of spread Methicillin-resistant Staphylococcus aureus (MRSA) and vancomycin-resistant Enterococcus (VRE) isolates by hospital day of admission. Early peak corresponds to patients entering the hospital with MRSA or VRE bacteremia. Later peak likely represents nosocomial acquisition. 55 (San Francisco County) Antibiotic resistance spreads to novel bacteria 56 www.mja.com.au Antibiotic use • Hospitals and communities create a metapopulation framework (Lipsitch et al; Smith et al) • Spatially- explicit strategies could help • Economics dominates control 57 Individuals may harbor ARB on admission…carriers • How do increases in the general population contribute to infections by ARB in the hospital, and what can be done about it? • Develop metapopulation models exploring colonization of hosts by antibiotic resistant strains 58 Individual movement Basic model structure k i indicates group, such as elderly j,k indicate subpopulations, such as hospital, community q indicates proportion (fixed) Model assumes admit=discharge 59 Smith et al, PNAS 2004 Bigger hospitals have bigger problems 60 Hospitals in larger cities have larger problems 61 Smith, Levin, Laxminarayan • Consider a game among hospitals • Compute optimal investment for a single hospital in controlling antibiotic resistance • Compute game-theoretic optimal strategy in a mixed population, with discounting • Investment decreases with city size 62 Conclusions • Infectious diseases have a rich modeling history • Great challenges for behavioral sciences • Relevant methods will span a broad range 63