1. Epidemiological determinants of spread of causal agent of severe acute respiratory syndrome in Hong Kong
This investigation into the epidemiology of SARS in Hong Kong was a collaboration between British and Chinese researchers. The intention was to identify critical epidemiological determinants of disease outcome. The identification of such factors could be helpful in improving management and the targeting of public health measures in an effective manner.
The investigators administered a questionnaire to all hospitalized SARS cases within 24 h of confirmation of the diagnosis. A total of 1425 cases, which included cases seen in Hong Kong up until April 28th 2003, were available for analysis. The key epidemiological parameters examined were: time from infection to symptom onset, time from onset to hospitalization, and time from admission to either discharge or death. The relationship of the estimated case fatality rate (CFR) to patient age and to the time from onset to admission was also assessed.
Significant findings were: a mean incubation period for the Hong Kong outbreak of 6.4 days (95% CI: 5.2–7.7 days), and a mean duration from symptom onset to hospital admission that varied from 3 to 5 days (with longer times prevailing earlier in the outbreak). The most striking findings related to age and the CFR. Using a parametric method of analysis the CFR for victims aged less than 60 years was 13.2% (95% CI: 9.8–16.8%), jumping to 43.3% (95% CI: 35.2–52.4%) for victims aged 60 or more. A non-parametric method of analysis yielded estimated CFRs of 6.8% (95% CI: 4.0–9.6%) for patients under 60, and 55% (95% CI: 45.3–64%) for those 60 and above. Time from onset of admission to hospitalization did not affect outcomes, although the authors make the point that early hospitalization and isolation remove a potential source of infection from the community.
This study reveals two striking findings: that the duration between symptom onset and hospital admission does not correlate with outcome, and that age is a potent predictor of poor outcome, with patients 60 and above faring particularly badly.
Lancet 361(9371):1761–1766
2. Audiometric changes associated with the treatment of uncomplicated falciparum malaria with co-artemether
The ototoxicity of quinine based and related antimalarials is well known. Concern has also been expressed in the past about the potential oto- and neuro-toxicity of the artemesinins, currently the most rapidly acting antimalarials available. While damage to brain stem nuclei has been seen in a wide range of experimental animals, including primates, administered parenteral artemesinins, oral artemesinins were thought free of such toxicity. The rationale behind this thinking was that the extremely short half life of orally administered artemisinins prevented their accumulation and any possibility of neurotoxicity.
The authors of this paper were able to analyze audiometric data collected for occupational health reasons at a construction site in Mozambique. At this construction site it was standard policy to treat all uncomplicated malaria with co-artemether (artemether-lumefantrine). All employees at the site underwent baseline audiometry before taking up duty, and had comparator audiograms taken on completion of their duties. Changes in hearing threshold for frequencies from 250 Hz to 8 kHz could then be derived for each frequency measured in left and right ears.
On completion of the study 150 subjects who had been treated with co-artemether were available for comparison with 150 age, gender, weight, and race matched controls. There was a statistically significant difference in the hearing threshold shifts at all frequencies measured in the range 1–8 kHz (p=0.001–0.04). The mean magnitude of the loss varied from 0.07 DB at 6 kHz–6.50 DB at 1 kHz in the treated group.
The mean interval between co-artemether exposure and the comparator audiogram was 164 days (range 3–392 days, SD 91.9 days). The authors suggest that this interval points to the loss being irreversible and question whether co-artemether may provoke other forms neurotoxicity.
The study did not determine which component of co-artemether might be responsible, whether both components might be synergistically responsible, or the mechanism of the toxicity. The safety of artemesinin combination therapy (ACT), especially in the pediatric setting may need to be reevaluated.
Transactions of the Royal Society of Tropical Medicine and Hygiene. 98(5): 261–267.
3. Transmission of the severe acute respiratory syndrome on aircraft
Given that air transportation of individuals infected with the SARS virus was instrumental in spreading the disease internationally, a valid question is whether transmission of the virus can occur on board aircraft. This study examined three separate flights that carried known individuals who met the WHO criteria for a probable SARS case.
The investigators set out to interview passengers and flight attendants on all three of these flights at least 10 days after flight completion. Fifty five percent of passengers and crew from the three flights were not available for interview; this total included four passengers from one flight who were reported to WHO as having developed SARS.
Transmission rates varied between flights, from no reported transmission on one flight of 90 min, to 22 confirmed or probable cases on another of 3 h duration. Of interest on the latter flight was that the transmission risk was greatest for passengers seated in the three rows in front of the index case: eight of 23 passengers in these rows came down with infection, as opposed to 10 of 88 passengers seated elsewhere (RR 3.1; 95% CI: 1.4–6.9). This particular finding indicates that SARS may be spread by aerosols and small droplets as well as large droplets, the infectivity range of large droplets being considered to not exceed 36 in. The fact that passengers seated fore of the index case were at greater risk points to coughing by the index patient as the means of spread.
The authors speculate on the reasons why their results show differing transmission rates on different flights, attributing this to variations in aircraft ventilation systems, aircraft type and size, flight duration, possible transmission prior to boarding, and the number of infected individuals on board. The authors also comment on the fact that they were able to interview 45% of passengers and crew, believing this to indicate how difficult contact tracing during an epidemic actually is.
The final conclusion of the study is that SARS, while unlikely to be transmitted aboard aircraft by asymptomatic individuals, may well be transmitted by individuals infected with SARS and displaying symptoms.
NEJM 349:2416–2422
4. The number needed to vaccinate (NNV) and population extensions of the NNV: comparison of influenza and pneumococcal vaccine 4 programmes for people aged 65 years and over
Australian investigators have developed a concept, similar to “number needed to treat” or “NNT” called “number needed to vaccinate” or NNV. This concept quantifies the number of doses of vaccine that need to be given annually to prevent a disease “incident”, defined in this study as a case of influenza, and of hospitalization or death due to influenza or invasive pneumococcal disease.
The cost per disease incident prevented can be calculated from the NNV, and the cost effectiveness of the vaccines in question, influenza and the polysaccharide pneumococcal vaccine, assessed. The investigators applied this technique to a population of Australians aged 65 years or more, and discovered that the cost each year of preventing a single case of influenza was $598, of preventing hospitalization due to influenza $10,787, and of preventing a single death attributable to influenza $74,801. Under the assumption that any case of invasive pneumococcal disease would be hospitalized, the cost per case of invasive pneumococcal disease prevented was $11,494; the cost per death prevented was $49,972.
Using this NNV analysis, the costs of preventing a single hospitalization attributable to influenza and the polysaccharide pneumococcal vaccine appeared to be almost the same. The authors make the point that influenza vaccination in the elderly has been shown in a number of studies to be cost effective. They then extrapolate to conclude that the nearly equal NNV findings for influenza and the pneumococcal polysaccharide vaccine indicate cost effectiveness of the latter vaccine in the age group under study.
The authors suggest that the likely impact and cost effectiveness of any vaccine programme could be measured using the same NNV tool. They also point out that findings should be considered specific to the target populations studied, as the impact of vaccine deployment will vary with the age, and other descriptors, of the populations studied. The overall conclusion though was that influenza and polysaccharide pneumococcal vaccination in Australians aged 65 years or more is a cost effective public health measure.
Vaccine. Article in press, available at www.sciencedirect.com
5. Thiomersal in vaccines: a regulatory perspective. WHO consultation
This paper records the discussions of a WHO convened consultative meeting on the use of thiomersal in vaccines. Represented were WHO, industry, regulators, and other interested parties. Concerns have been raised over the use of the organic mercurial thiomersal as a preservative and antibacterial in vaccines. The FDA in the United States has mandated that vaccines for use in that country should eventually contain no, or the minimum possible amount, of thiomersal. The removal or reduction of thiomersal is not a simple matter, as the substance is used not only as a preservative (especially important when multi-dose vials are used) but also as an inhibitor of various processes during vaccine manufacture.
For this reason the removal of thiomersal from vaccines may lead to them becoming more reactogenic, as happened when the substance was eliminated from a European tick borne encephalitis (TBE) vaccine. It appears that thiomersal inhibits generation of the cytokines IL-1, IL-6, and TNFα, and removal of thiomersal from the TBE vaccine rendered it more pyrogenic. This finding has led to the need to develop new tests for pyrogenicity based upon cytokine production in human blood rather than animal testing. In the case of an rDNA hepatitis B vaccine, an approximately 96% reduction in thiomersal content led to greater HBsAg content in the vaccine, and increased immunogenicity.
Concerns have been raised that the elimination of thiomersal from the type of multidose vials used in vaccination campaigns in developing countries may lead to contamination, and it appears as if thiomersal will be retained as a preservative in such situations. In developed countries the move will be towards thiomersal free, or nearly free, single dose ampoules or pre-filled syringes.
The main recommendations of the WHO consultative process were that the elimination of thiomersal from vaccines was not a straightforward process, as illustrated by the examples above; that calls for its removal may be driven by public pressure and not by good science; that limits for human intake of ethyl mercury derivatives such as thiomersal should not be extrapolated from those for methyl mercury derivatives, and as a general principle only additives shown essential on scientific grounds should be added to vaccines.
Vaccine. Article in press, available at www.sciencedirect.com
6. The meningococcal vaccine—public policy and individual choices
This paper examines, in the light of the availability of meningococcal vaccines, the tension between public health policy on vaccine recommendation and the rights of individuals to exercise free choice.
The authors recount that meningococcal disease has a high case fatality rate (10%), and that infection may lead to serious neurological sequelae and limb amputation. The paper focuses on the current polysaccharide meningococcal vaccines active against serogroups A, C, Y, and W-135, which account for the majority of cases worldwide. Production of a polysaccharide vaccine against serogroup B has been prevented by the similarity of the meningococcal sialic Acid antigen to the polysialic glycopeptides of a human neural-cell adhesion molecule.
In the United States the incidence of disease caused by the various serogroups varies by geographic location and other factors, but serogroup B is the most important in children aged 1 year or less. Serogroup A is the commonest serogroup encountered in sub-Saharan Africa, a region prone to outbreaks and epidemics.
The incidence of meningococcal disease falls with age in the US population, dropping from 7.1 per 100,000 in children aged 1 year or less, to 0.7 per 100,000 in individuals aged between 5 and 34 years. The poor immunogenicity of the polysaccharide vaccine in the very young and the importance of serogroup B in this age group render use of the polysaccharide vaccine to the children at highest risk pointless. The reduced incidence in older children and adolescents, in whom the vaccine is immunogenic, makes the vaccine not cost effective from a public health view point however. Calculations are that to prevent a single case of meningococcal disease would cost $4 m, while preventing a single death would cost $48 m. While this places the vaccine beyond the reach of the public purse, the vaccine is still available to those who wish to protect themselves or their children against a much feared disease.
The problem the authors identify is a low awareness and acceptance of vaccines that are not recommended as part of routine immunization programmes. This lack of awareness extends both within the lay population and the medical profession, a common misperception being that if a vaccine is not included in routine immunization programmes, then it is either ineffective or unsafe.
The authors recommend that policies be put into place to create a greater awareness of the meningococcal vaccine, which they say “could save lives and prevent the devastating consequences of meningococcal infection.”
NEJM 349:2353–2356
7. Frequency of venous thromboembolism in low to moderate risk long distance air travellers: the New Zealand Air Traveller's Thrombosis (NZATT) study
This New Zealand based study utilized the geographical isolation of that country to scientific advantage, examining the incidence of deep vein thrombosis and pulmonary embolism in long haul travelers.
This was a large prospective study, with 878 subjects ultimately recruited and available for analysis. Subjects were New Zealand residents in the age range 18–70 years. A baseline assessment of putative risk factors was obtained before departure, along with a pre-departure D-dimer assay.
Subjects excluded were those with raised baseline D-dimers or a history of venous thromboembolism, those taking anticoagulants, individuals with malignant disease, those planning to be away from New Zealand for more than 6 weeks, and pregnant women. All subjects flew at least 10 h, with an average time airborne during absence from New Zealand of 39 h (attributable to multiple flight itineraries). On return all subjects completed a questionnaire detailing their itineraries and in-flight behavior, including the use of prophylactic measures such as the wearing of compression stockings and aspirin. A clinical assessment was also performed upon return and D-dimer levels measured. D-dimer measurement was repeated a further 2 weeks after return. Patients with raised D-dimers on return were further investigated with bilateral compression ultrasonography and computerized tomographic pulmonary angiography. A total of 112 subjects underwent imaging, which revealed four cases of pulmonary embolus and five of deep vein thrombosis; six of these patients had pre-existing risk factors for thromboembolism, two were discovered post travel to have a thrombophilic coagulation profie, and one was diagnosed with carcinoma of the breast 2 months after her return.
Of these nine subjects with radiologically proven thromboembolism, four had worn compression stockings, and five had taken aspirin. The study did deliberately not seek to influence prophylactic behaviors, and did not have a control group that could inform on the background venous thromboembolism rate in the study population. The incidence of thromboembolism while abroad could also not be assessed, leading to the final incidence quoted of 1% (9/878) being a conservative one.
Thromboembolic events occurred in all classes of travel, emphasizing the wisdom of dropping the term ‘economy class syndrome’ in favor of “traveler's thrombosis” or “air travel related thrombosis”. The findings show a clear relation between long haul air travel (>10 h) and venous thromboembolism, especially in at risk individuals.
Lancet 362:2039–2044
8. Formation of edema and fluid shifts during long haul flights
This study was undertaken to assess the degree of edema that might develop in passengers during long haul flights, and to determine whether there was any difference in this regard between passengers with and without known risks for deep vein thrombosis (DVT). It is the first such study to have been undertaken under real flight conditions.
The study employed plethysmography to demonstrate and track the development of dependent edema in passengers. Ten volunteers with known risk factors for DVT (mean body mass index 33.4) and 10 volunteers without significant risk factors (mean body mass index 23.6) were flown from Innsbruck to Vienna (1 h) and then on to Washington (8 h 20 min). Subjects were permitted to remain in Washington for 2 days and then returned to Innsbruck. On board fluid consumption was monitored and alcohol restricted to 0.5 l of beer each way.
Prior to departure measurements of pretibial and forehead skin thickness were taken, along with volume measurements of the legs by plethysmography. These measurements were repeated on board between the fifth and eighth hours of flight in both directions, immediately after landing in Vienna, and again 1 and 3 days after return to Vienna.
Plethysmography was undertaken using an optico-electronic scanner (Perometer), and skin thickness was measured using ultrasonic pulse echo equipment (CL3DL, Krautkraemer and Co). Additionally, hemoglobin level and packed cell volumes were measured along with each plethysmographic measurement.
There were no statistically significant differences between the results of the group with risk factors and the group without, and none of the 20 volunteers was found to be suffering from a DVT as assessed on return by compression duplex sonography. Mean leg volume increase between the pre-departure and the immediate post-return measurements was 254 ml in each leg. Pretibial skin thickness increased significantly during the flight in comparison with pre-departure measurements. Pretibial thickness remained elevated the day following return, returning to normal when reassessed 3 day post-return. Interestingly, forehead skin thickness was found to increase above baseline levels on the return but not on the outward flight, and to drop below baseline measurements 3 days after return.
Hemoglobin levels were significantly (p<0.001) reduced on return (mean 145 g/l, range 125–148 g/l) when compared to pre-departure measurements (mean 148.5 g/l, range 137–155 g/l). Pack cell volumes showed a similar pattern with pre-departure levels (mean 44.5%, range 41–47%) significantly higher (p<0.01) than return flight levels (mean 43.5%, range 37–45%). These decreases are ascribed by the authors to in-flight fluid consumption.
The authors conclude that fluid shifts during flight are significant, occur rapidly, and persist post-flight, and that these shifts provide a good argument in support of the prescription of graduated pressure stockings to travellers.
J Travel Med 10:334–339.