Wednesday, March 20, 2013

Blogging CDC 2013 Canberra, Australia, Day 1

Blogging some highlights from the Australian CDC 2013 Conference - the peak biannual communicable disease surveillance conference in Australia. The presentations should be available on the website soon. Some highlights from the morning panel session was John Kaldor's (from Kirby Institute) comment that the decreased rate of viral warts in young men prior to the roll out of the male HPV vaccine suggests "that we probably don't need to vaccinate guys as the women have it covered".   Rosemary Lester, CHO of Victoria - but speaking as chair of the Communicable Disease Network of Australia described a vision of national surveillance out to 2020. Allan Cheng from University of Melbourne asked what were the barriers to a CDC i.e. a federal centre for disease control akin to the US CDC or European ECDC in Australia.  It was obvious that this was an uncomfortable question for some of the panelists.More highlights..
Such is the sensitivity surrounding the implementation of an Australian CDC that one of the commentatators referred to is as the "C" word.  For those not familiar with the Australian system - surveillance and outbreak investigation is primarily the responsibility of the states and territories.  At the Commonwealth level there is a Communicable Disease Network of Austraia (CDNA) committee that coordinate much of the surveillance and outbreak investigation across the country but many of the enhanced surveillance and public health laboratory systems are independent programs or centres with national funding (and sometimes state funding).  The need for an Australian CDC was the subject of a hot debate at the 2011 CDC conference. 
Mahomed Patel from ANU presented on the achievements of the Master of Applied Epidemiology Program - the Australian FETP akin to the CDC EIS program (I was a supervisor on the program from 1995-97).  Mahomed then revisited a 1987 MJA article by Bob Douglas raising the need for a national CDC in Australia - obviously  a long debate. Reflecting on the important positions held by MAE graduates in the wider communicable disease network and some of the important outbreak and surveillance response they have contributed to is evidence of the value of the program - some statistics: since 1991, 167 graduates, > 400 outbreak investigations, established 30 surveillance systems, and published 450 papers.  He noted that MAE graduate Martyn Kirk from ANU was the convenor of the CDC 2013 and the first coordinator of the national OzFoodnet foodborne surveillance network.

Mahomed's talk then turned to system issues (my ears pricked up), quoting Peter Senges work on Integrated Learning and then went on to present an interesting table on characteristics of resilient networks see page 251 in Marsh and Rhodes Policy Networks in British Government Oxford  OUP1992.

Stephen Lambert from AID presented "Is Communicable Disease Surveillance Broken and Can We Fit It?  He  focused on difficulties in interpreting notifiable disease surveillance in particular with regard to selected vaccine preventable diseases.  He discussed influenza - biased by number of tests performed - Queensland performing higher numbers of tests and therefore appearing to have higher rates of disease but same positivity as other jurisdictions.  He makes the point that it is difficult to compare influenza laboratory notificaiton rates between jurisdictions. (Thats why we have Flutracking - standard method of ILI surveillance across the country.) He also highlighted how the frequency of testing and testing sensitivity methods drove much of the surveillance data reported for rotavirus and pertussis.  Stephen suggested the solution was to mandate reporting of all laboratory testing i.e. negative results as well as well as positive.   During the question session Rob Hall suggested that the fraction of cases detected by surveillance was an important parameter that should be captured, Allen Cheng suggested sentinel data could provide insight into  biases into the notified data and I suggested that we could adjust the surveillance data similar to the way economists adjust housing prices to a base year to take into account inflation.  e.g. adjust up influenza notification for years prior to the availability of PCR testing to provide a more realistic estimate of cases. Stephen was not very keen on these suggestions and saw them as distractions from the main game (he described my suggestion as "lazy" - I was suggesting adjustment for historical purposes in addition to mandatory notification of testing). However, these other suggestion probably have their place - they may help us to understand changes in testing and laboratory practices that can distort the surveillance data.

Now to the afternoon influenza stream. Clayton Chiu from NCIR presesnted data on the severity of influenza among children based on hospital discharge data and suggested a universal influenza vaccination program for children.  Louise Maher explored predictors of influenza vaccination during pregancy - women who received a health care provider recommendation for influenza vaccination were 40 times more likely to receive the vaccine. Only 10% did not get the vaccine due to safety concerns. Holly Seale from UNSW explored attitudes to using surgical and other masks for protection of health care workers in hospitals. Most frequent reasons for not wearing include: uncomfortable, poorly fitted, interfere with communication, undesired alarming impressions. Allen Cheng presented influenza vaccine effectiveness data from Flucan finding that the flu vaccine provided a 40% reduction in hospitalisation for influenza related illness and was less protective among adults > 65 years of age. Ee Laine Tay analysed Victorian influenza surveillance data using the new WHO "baseline", "average" and "above average" threshold methods. Frances Birrell reported that the median duration of a nursing home influenza outbreak was 14 days with a median attack rate of 12.4% -higher in residents than staff. Jodie McVernon explored antibody prevalance to the new variant H3N2 swine influenza strain that emerged in the USA in 2012.  Interestingly there was a cohort effect showing there was cross reactivity to a similar influenza strain in subjects 16 years plus  - i.e. under 10 years of age group likely susceptible to the novel H3N2 strain should it become easily person-to-person spread (apparently need to literally hug a pig to acquire it thus far). Donna Mak used data linkage for > 25% identification as Aboriginal or Torres Strait Islander to improved identification.

No comments: