Correspondence
scarcity of primary data from lowincome countries, especially subSaharan Africa. For funding bodies and policy makers who use these estimates to guide allocation of scarce health resources, HAI and AMR in lowincome countries are therefore not just neglected diseases, but invisible ones. A large study of paediatric admissions at one hospital in rural Kenya between 2002 and 2009 estimated nosocomial bacteraemia to occur in six of 1000 admissions, with a very high associated mortality (53%).3 Deaths in this study were attributable, at least partly, to antimicrobial resistance. Based on these data, we estimated the disease-specific burden in sub-Saharan African children in 2005 by applying the admission rates,4 risk per admission, case-fatality ratio, and impact per case3 determined at this site for regional child population estimates.5 On the basis of these assessments, nosocomial bacteraemia could have accounted for 25 000 deaths and 270 000 additional inpatient days in African children in 2005. This approach has limitations; essentially, we multiplied the disease rate at a single site by the African child population, nonetheless, HAI and AMR contribute substantially to the morbidity and mortality burden in African children. Furthermore, detectable nosocomial bloodstream infections are merely the tip of the iceberg; most episodes of HAI are not bacteraemic, and the sensitivity of paediatric blood cultures is inherently poor. Results from studies in high-income countries show that HAI are usually preventable, and interventions are usually cost effective to implement and sustain. Health-care facilities in low-income countries face different challenges to infection control to those in high-income settings2— in low-income facilities, the effect of introducing infection control interventions largely remains to be determined. However, in view of the large size of this problem, and the affordability of interventions such as improved hand hygiene and surgical www.thelancet.com/infection Vol 14 July 2014
checklists, the gains could be high and hugely cost effective. A reduction in the amount of infections caused by antibiotic-resistant bacteria at their source would undoubtedly have measurable local and global health benefits. We declare no competing interests.
*Alexander M Aiken, Benedetta Allegranzi, J Anthony Scott, Shaheen Mehtar, Didier Pittet, Hajo Grundmann [email protected] London School of Hygiene and Tropical Medicine, WC1E 7HT, London, UK (AMA, JAS); Service Delivery and Safety, WHO, Geneva, Switzerland (BA); KEMRIWellcome Trust Research Programme, Kilifi, Kenya (JAS); Unit for Infection Prevention and Control, Division of Community Health, Faculty of Health Sciences, Stellenbosch University, Cape Town, South Africa (SM); Infection Control Program and WHO Collaborating Centre on Patient Safety, University of Geneva Hospitals and Faculty of Medicine, Geneva, Switzerland (DP); and Department of Medical Microbiology, University of Groningen, University Medical Centre, Groningen, Netherlands (HG) 1
2
3
4
5
Laxminarayan R, Duse A, Wattal C, et al. Antibiotic resistance—the need for global solutions. Lancet Infect Dis 2013; 13: 1057–98. Allegranzi B, Bagheri Nejad S, Combescure C, et al. Burden of endemic health-careassociated infection in developing countries: systematic review and meta-analysis. Lancet 2011; 377: 228–41. Aiken AM, Mturi N, Njuguna P, et al. Risk and causes of paediatric hospital-acquired bacteraemia in Kilifi District Hospital, Kenya: a prospective cohort study. Lancet 2011; 378: 2021–27. Moisi JC, Gatakaa H, Berkley JA, et al. Excess child mortality after discharge from hospital in Kilifi, Kenya: a retrospective cohort analysis. Bull World Health Organ 2011; 89: 725–32. United Nations, Department of Economic and Social Affairs, Population Division, Population Estimates and Projections Section. World population prospects: the 2012 revision. http://esa.un.org/unpd/wpp/unpp/panel_ indicators.htm (accessed Feb 26, 2014).
Assessing the pandemic potential of emerging influenza Accurate identification of patients’ demo graphic charac teristics and case counts are key factors in understanding the pandemic potential of emerging influenza viruses, and the implementation of effective medical
surveillance and public health response actions. A recent Editorial in The Lancet Infectious Diseases1 contained an error about the provenance of the first human being with H5N1 avian influenza reported in the western hemisphere, and omitted key data from recent cases of H10N8 avian influenza in human beings in China. The first confirmed case of H5N1 detected in a human being in the Americas was in a woman, not a man—a hospital health-care worker who died around 1 week after her return to Canada from Beijing, China. The woman, who was travelling with a family member, experienced the first onset of symptoms during her return flight from Beijing to Canada on Dec 27, 2013. She was admitted to hospital on Jan 1, 2014, and died 2 days later on Jan 3.2,3 Local media reported that the woman was in her 20s, the age group at highest risk of death from H5N1.4 The Editorial cited a fatal case of H10N8 in December, 2013, in a man aged 75 years from China’s Jiangxi Province. However, two more cases, including a fatality, were reported from China’s Jiangxi Province between Jan 29, and Feb 13, 2013. The first additional case that ended in a fatality was in a woman aged 73 years, and the second case was a 55-year-old woman.5,6 These cases were detected through intensive ongoing surveillance for H7N9 cases in China, which have been most common in elderly people.4 The emergence and spread of new viruses should be tracked, and their pandemic potential predicted and countered more readily by the early identification of key, at-risk populations. The high variability of age and sex differences in morbidity and mortality rates from emerging zoonotic viruses, such as new coronaviruses and influenza viruses, can provide important indices of populations at highest risk of infection and death, and insights into potential mechanisms for transmission of these viruses to human beings, and between people. This information, 551
Correspondence
Published Online June 12, 2014 http://dx.doi.org/10.1016/ S1473-3099(14)70807-2 See Articles page 619
in turn, can help to develop public health strategies for the most efficient and effective use of scarce diagnostic, medical care, and treatment resources. I declare no competing interests.
Joseph P Dudley [email protected] Leidos Corporation, 12530 Parklawn Drive, Suite 350, Rockville, MD 20855, USA 1
2
3
4
5
6
The Lancet Infectious Diseases. Pandemic potential of emerging influenza. Lancet Infect Dis 2014; 14: 173. European Centre for Disease Control and Prevention. Epidemiological update: avian influenza A(H5N1) Jan 9, 2014. http://ecdc. europa.eu/en/press/news/_layouts/forms/ News_DispForm.aspx?List=8db7286c-fe2d476c-9133-18ff4cb1b568&ID=938 (accessed Feb 28, 2014). Pabbaraju K, Tellier R, Wong S, et al. Full-genome analysis of avian influenza A(H5N1) virus from human, North America, 2013. Emerg Infect Dis 2014; 20: 887–91. Dudley JP, Mackay IM. Age-specific and sex-specific morbidity and mortality from avian influenza A(H7N9). J Clin Virol 2013; 58: 568–70. WHO-Western Pacific Regional Office. Avian influenza A (H10N8): update as of January 30, 2014. Manila. http://www.wpro.who.int/china/ mediacentre/factsheets/h10n8/en/ (accessed Feb 28, 2014). Hong Kong Department of Health, Centre for Health Protection. Notification by the China National Health and Family Planning Commission of a fatal human case of avian influenza A(H10N8) in Jiangxi. Feb 14, 2014. http://www.chp.gov.hk/en/content/ 116/33426.html (accessed Feb 28, 2014).
Dengue outlook for the World Cup in Brazil Because the 2014 FIFA World Cup in Brazil is approaching soon, estimation of the dengue risk for this period in Brazil is important. We therefore commend Rachel Lowe and colleagues1 on their Article in which they discuss how they developed an early warning system for dengue, based on a spatiotemporal Bayesian hierarchical model framework driven by climate and non-climate information.1 They identified optimum trigger alert thresholds for scenarios of mediumrisk and high-risk of dengue, thus enabling public health practitioners to implement early interventions. The paper correctly concludes that there is a higher risk of dengue in the cities of Fortaleza and Natal during the time of the World Cup. However, we disagree with the order of magnitude described in their paper. We calculated the risk of dengue for foreign visitors to the World Cup on the basis of past daily (not monthly) incidence in the 12 cities that will host the games.2 Our calculations also included estimates of the expected number of visitors in each city, and the
Lowe et al estimations1 Expected plow number of visitors*
Massad et al estimations2
pmedium
phigh
Minimum expected number of dengue cases†
Expected number of dengue cases‡
Expected number of dengue cases per 100 000 visitors*‡
Natal
44 952
32%
20%
48%
>74
6 (2–9)
13·3 (4·4–20·0)
Fortaleza
58 195
34%
20%
46%
>92
10 (1–14)
17·2 (1·7–24·1)
Rio de Janeiro
101 910
62%
25%
13%
>66
11 (0–22)
10·8 (0·0–21·6)
Belo Horizonte
36 788
65%
24%
11%
>21
3 (0–8)
8.2 (0·0–21·7)
Recife
37 693
57%
24%
19%
>31
0 (0–1)
0·0 (0·0–2·7)
Cuiabá
20 740
71%
22%
7%
>9
2 (0–4)
9·6 (0·0–19·3)
Brasília
34 391
73%
20%
7%
>14
1 (0–1)
2·9 (0·0–2·9)
Salvador
57 855
56%
27%
17%
>45
0
0·0
Manaus
26 252
63%
25%
12%
>16
0
0·0
101 150
99%
1%
0%
>2
0
0·0
Curitiba
34 782
100%
0%
0%
0
0
0·0
Porto Alegre
35 343
100%
0%
0%
0
0
0·0
33 (3–59)
5·4 (0·5–9·7)
São Paulo
Total
607 051
>370
plow=incidence lower than 100 per 100 000. pmedium=incidence between 100 and 300 per 100 000. phigh=incidence higher than 300 per 100 000. *Source: EMBRATUR. †Minimum number of expected cases (weighted mean)>1 × plow + 100 × pmedium + 300 × phigh. ‡Average (best–worst) scenarios.
Table: Expected number of visitors and dengue cases for each of the Brazilian cities hosting the World Cup games
552
expected number of days that visitors will spend in each city, depending on each of the 32 possible schedules of the games. By multiplying these individual risks by the proportion of the expected number of visitors with respect to each stadium capacity, we estimated the expected number of dengue cases in each of these 12 cities. Our estimations are, on average, more than ten-times lower than those from Lowe and colleagues (table). We estimated the minimum expected number of dengue cases by taking the weighted mean considering the lower bounds of Lowe and colleagues’ estimated incidences (plow, pmedium, phigh). We estimate 33 cases (range 3–59) among the projected 600 000 visitors. The strength of our analysis is that it includes correlations with the exact FIFA match schedule, and best-toworst case scenarios. Our results support a low dengue risk for visitors to the World Cup, which is consistent with data recently published by GeoSentinel, a network of travel medicine providers.3,4 Predictions are still merely predictions and depend on assumptions from past experiences. Of course, the incidence of dengue in 2014 might be ten-times higher than ever; it is just unlikely to be so (in fact, the number of cases reported so far in 2014 are 43% fewer in Natal and 35% fewer in Fortaleza, than in 2013). However, health-care providers in countries where World Cup visitors will return should be at high alert for dengue, and report cases immediately to authorities. In doing so, timely surveillance can be established and provide the true number of dengue cases during the World Cup. The research was partially funded by LIM01HCFMUSP, FAPESP, CNPq (EM, MNB, RX, MA), the Brazilian Ministry of Health (MNB), and Dengue Tools under the Seventh Framework Programme of the European Community (EM, AWS). We declare no competing interests.
Eduardo Massad, Marcelo N Burattini, Raphael Ximenes, Marcos Amaku, *Annelies Wilder-Smith [email protected] www.thelancet.com/infection Vol 14 July 2014
scarcity of primary data from lowincome countries, especially subSaharan Africa. For funding bodies and policy makers who use these estimates to guide allocation of scarce health resources, HAI and AMR in lowincome countries are therefore not just neglected diseases, but invisible ones. A large study of paediatric admissions at one hospital in rural Kenya between 2002 and 2009 estimated nosocomial bacteraemia to occur in six of 1000 admissions, with a very high associated mortality (53%).3 Deaths in this study were attributable, at least partly, to antimicrobial resistance. Based on these data, we estimated the disease-specific burden in sub-Saharan African children in 2005 by applying the admission rates,4 risk per admission, case-fatality ratio, and impact per case3 determined at this site for regional child population estimates.5 On the basis of these assessments, nosocomial bacteraemia could have accounted for 25 000 deaths and 270 000 additional inpatient days in African children in 2005. This approach has limitations; essentially, we multiplied the disease rate at a single site by the African child population, nonetheless, HAI and AMR contribute substantially to the morbidity and mortality burden in African children. Furthermore, detectable nosocomial bloodstream infections are merely the tip of the iceberg; most episodes of HAI are not bacteraemic, and the sensitivity of paediatric blood cultures is inherently poor. Results from studies in high-income countries show that HAI are usually preventable, and interventions are usually cost effective to implement and sustain. Health-care facilities in low-income countries face different challenges to infection control to those in high-income settings2— in low-income facilities, the effect of introducing infection control interventions largely remains to be determined. However, in view of the large size of this problem, and the affordability of interventions such as improved hand hygiene and surgical www.thelancet.com/infection Vol 14 July 2014
checklists, the gains could be high and hugely cost effective. A reduction in the amount of infections caused by antibiotic-resistant bacteria at their source would undoubtedly have measurable local and global health benefits. We declare no competing interests.
*Alexander M Aiken, Benedetta Allegranzi, J Anthony Scott, Shaheen Mehtar, Didier Pittet, Hajo Grundmann [email protected] London School of Hygiene and Tropical Medicine, WC1E 7HT, London, UK (AMA, JAS); Service Delivery and Safety, WHO, Geneva, Switzerland (BA); KEMRIWellcome Trust Research Programme, Kilifi, Kenya (JAS); Unit for Infection Prevention and Control, Division of Community Health, Faculty of Health Sciences, Stellenbosch University, Cape Town, South Africa (SM); Infection Control Program and WHO Collaborating Centre on Patient Safety, University of Geneva Hospitals and Faculty of Medicine, Geneva, Switzerland (DP); and Department of Medical Microbiology, University of Groningen, University Medical Centre, Groningen, Netherlands (HG) 1
2
3
4
5
Laxminarayan R, Duse A, Wattal C, et al. Antibiotic resistance—the need for global solutions. Lancet Infect Dis 2013; 13: 1057–98. Allegranzi B, Bagheri Nejad S, Combescure C, et al. Burden of endemic health-careassociated infection in developing countries: systematic review and meta-analysis. Lancet 2011; 377: 228–41. Aiken AM, Mturi N, Njuguna P, et al. Risk and causes of paediatric hospital-acquired bacteraemia in Kilifi District Hospital, Kenya: a prospective cohort study. Lancet 2011; 378: 2021–27. Moisi JC, Gatakaa H, Berkley JA, et al. Excess child mortality after discharge from hospital in Kilifi, Kenya: a retrospective cohort analysis. Bull World Health Organ 2011; 89: 725–32. United Nations, Department of Economic and Social Affairs, Population Division, Population Estimates and Projections Section. World population prospects: the 2012 revision. http://esa.un.org/unpd/wpp/unpp/panel_ indicators.htm (accessed Feb 26, 2014).
Assessing the pandemic potential of emerging influenza Accurate identification of patients’ demo graphic charac teristics and case counts are key factors in understanding the pandemic potential of emerging influenza viruses, and the implementation of effective medical
surveillance and public health response actions. A recent Editorial in The Lancet Infectious Diseases1 contained an error about the provenance of the first human being with H5N1 avian influenza reported in the western hemisphere, and omitted key data from recent cases of H10N8 avian influenza in human beings in China. The first confirmed case of H5N1 detected in a human being in the Americas was in a woman, not a man—a hospital health-care worker who died around 1 week after her return to Canada from Beijing, China. The woman, who was travelling with a family member, experienced the first onset of symptoms during her return flight from Beijing to Canada on Dec 27, 2013. She was admitted to hospital on Jan 1, 2014, and died 2 days later on Jan 3.2,3 Local media reported that the woman was in her 20s, the age group at highest risk of death from H5N1.4 The Editorial cited a fatal case of H10N8 in December, 2013, in a man aged 75 years from China’s Jiangxi Province. However, two more cases, including a fatality, were reported from China’s Jiangxi Province between Jan 29, and Feb 13, 2013. The first additional case that ended in a fatality was in a woman aged 73 years, and the second case was a 55-year-old woman.5,6 These cases were detected through intensive ongoing surveillance for H7N9 cases in China, which have been most common in elderly people.4 The emergence and spread of new viruses should be tracked, and their pandemic potential predicted and countered more readily by the early identification of key, at-risk populations. The high variability of age and sex differences in morbidity and mortality rates from emerging zoonotic viruses, such as new coronaviruses and influenza viruses, can provide important indices of populations at highest risk of infection and death, and insights into potential mechanisms for transmission of these viruses to human beings, and between people. This information, 551
Correspondence
Published Online June 12, 2014 http://dx.doi.org/10.1016/ S1473-3099(14)70807-2 See Articles page 619
in turn, can help to develop public health strategies for the most efficient and effective use of scarce diagnostic, medical care, and treatment resources. I declare no competing interests.
Joseph P Dudley [email protected] Leidos Corporation, 12530 Parklawn Drive, Suite 350, Rockville, MD 20855, USA 1
2
3
4
5
6
The Lancet Infectious Diseases. Pandemic potential of emerging influenza. Lancet Infect Dis 2014; 14: 173. European Centre for Disease Control and Prevention. Epidemiological update: avian influenza A(H5N1) Jan 9, 2014. http://ecdc. europa.eu/en/press/news/_layouts/forms/ News_DispForm.aspx?List=8db7286c-fe2d476c-9133-18ff4cb1b568&ID=938 (accessed Feb 28, 2014). Pabbaraju K, Tellier R, Wong S, et al. Full-genome analysis of avian influenza A(H5N1) virus from human, North America, 2013. Emerg Infect Dis 2014; 20: 887–91. Dudley JP, Mackay IM. Age-specific and sex-specific morbidity and mortality from avian influenza A(H7N9). J Clin Virol 2013; 58: 568–70. WHO-Western Pacific Regional Office. Avian influenza A (H10N8): update as of January 30, 2014. Manila. http://www.wpro.who.int/china/ mediacentre/factsheets/h10n8/en/ (accessed Feb 28, 2014). Hong Kong Department of Health, Centre for Health Protection. Notification by the China National Health and Family Planning Commission of a fatal human case of avian influenza A(H10N8) in Jiangxi. Feb 14, 2014. http://www.chp.gov.hk/en/content/ 116/33426.html (accessed Feb 28, 2014).
Dengue outlook for the World Cup in Brazil Because the 2014 FIFA World Cup in Brazil is approaching soon, estimation of the dengue risk for this period in Brazil is important. We therefore commend Rachel Lowe and colleagues1 on their Article in which they discuss how they developed an early warning system for dengue, based on a spatiotemporal Bayesian hierarchical model framework driven by climate and non-climate information.1 They identified optimum trigger alert thresholds for scenarios of mediumrisk and high-risk of dengue, thus enabling public health practitioners to implement early interventions. The paper correctly concludes that there is a higher risk of dengue in the cities of Fortaleza and Natal during the time of the World Cup. However, we disagree with the order of magnitude described in their paper. We calculated the risk of dengue for foreign visitors to the World Cup on the basis of past daily (not monthly) incidence in the 12 cities that will host the games.2 Our calculations also included estimates of the expected number of visitors in each city, and the
Lowe et al estimations1 Expected plow number of visitors*
Massad et al estimations2
pmedium
phigh
Minimum expected number of dengue cases†
Expected number of dengue cases‡
Expected number of dengue cases per 100 000 visitors*‡
Natal
44 952
32%
20%
48%
>74
6 (2–9)
13·3 (4·4–20·0)
Fortaleza
58 195
34%
20%
46%
>92
10 (1–14)
17·2 (1·7–24·1)
Rio de Janeiro
101 910
62%
25%
13%
>66
11 (0–22)
10·8 (0·0–21·6)
Belo Horizonte
36 788
65%
24%
11%
>21
3 (0–8)
8.2 (0·0–21·7)
Recife
37 693
57%
24%
19%
>31
0 (0–1)
0·0 (0·0–2·7)
Cuiabá
20 740
71%
22%
7%
>9
2 (0–4)
9·6 (0·0–19·3)
Brasília
34 391
73%
20%
7%
>14
1 (0–1)
2·9 (0·0–2·9)
Salvador
57 855
56%
27%
17%
>45
0
0·0
Manaus
26 252
63%
25%
12%
>16
0
0·0
101 150
99%
1%
0%
>2
0
0·0
Curitiba
34 782
100%
0%
0%
0
0
0·0
Porto Alegre
35 343
100%
0%
0%
0
0
0·0
33 (3–59)
5·4 (0·5–9·7)
São Paulo
Total
607 051
>370
plow=incidence lower than 100 per 100 000. pmedium=incidence between 100 and 300 per 100 000. phigh=incidence higher than 300 per 100 000. *Source: EMBRATUR. †Minimum number of expected cases (weighted mean)>1 × plow + 100 × pmedium + 300 × phigh. ‡Average (best–worst) scenarios.
Table: Expected number of visitors and dengue cases for each of the Brazilian cities hosting the World Cup games
552
expected number of days that visitors will spend in each city, depending on each of the 32 possible schedules of the games. By multiplying these individual risks by the proportion of the expected number of visitors with respect to each stadium capacity, we estimated the expected number of dengue cases in each of these 12 cities. Our estimations are, on average, more than ten-times lower than those from Lowe and colleagues (table). We estimated the minimum expected number of dengue cases by taking the weighted mean considering the lower bounds of Lowe and colleagues’ estimated incidences (plow, pmedium, phigh). We estimate 33 cases (range 3–59) among the projected 600 000 visitors. The strength of our analysis is that it includes correlations with the exact FIFA match schedule, and best-toworst case scenarios. Our results support a low dengue risk for visitors to the World Cup, which is consistent with data recently published by GeoSentinel, a network of travel medicine providers.3,4 Predictions are still merely predictions and depend on assumptions from past experiences. Of course, the incidence of dengue in 2014 might be ten-times higher than ever; it is just unlikely to be so (in fact, the number of cases reported so far in 2014 are 43% fewer in Natal and 35% fewer in Fortaleza, than in 2013). However, health-care providers in countries where World Cup visitors will return should be at high alert for dengue, and report cases immediately to authorities. In doing so, timely surveillance can be established and provide the true number of dengue cases during the World Cup. The research was partially funded by LIM01HCFMUSP, FAPESP, CNPq (EM, MNB, RX, MA), the Brazilian Ministry of Health (MNB), and Dengue Tools under the Seventh Framework Programme of the European Community (EM, AWS). We declare no competing interests.
Eduardo Massad, Marcelo N Burattini, Raphael Ximenes, Marcos Amaku, *Annelies Wilder-Smith [email protected] www.thelancet.com/infection Vol 14 July 2014
