If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
In November 2010, following reports of an outbreak of a fatal, febrile, hemorrhagic illness in northern Uganda, the Uganda Ministry of Health established multisector teams to respond to the outbreak.
Methods
This was a case-series investigation in which the response teams conducted epidemiological and laboratory investigations on suspect cases. The cases identified were line-listed and a data analysis was undertaken regularly to guide the outbreak response.
Results
Overall, 181 cases met the yellow fever (YF) suspected case definition; there were 45 deaths (case fatality rate 24.9%). Only 13 (7.5%) of the suspected YF cases were laboratory confirmed, and molecular sequencing revealed 92% homology to the YF virus strain Couma (Ethiopia), East African genotype. Suspected YF cases had fever (100%) and unexplained bleeding (97.8%), but jaundice was rare (11.6%). The overall attack rate was 13 cases/100 000 population, and the attack rate was higher for males than females and increased with age. The index clusters were linked to economic activities undertaken by males around forests.
Conclusions
This was the largest YF outbreak ever reported in Uganda. The wide geographical case dispersion as well as the male and older age preponderance suggests transmission during the outbreak was largely sylvatic and related to occupational activities around forests.
Yellow fever (YF) is an acute viral hemorrhagic disease caused by the yellow fever virus. The YF virus is an enveloped positive sense single-stranded RNA virus that belongs to the genus Flavivirus.
in: Zuckerman A.J. Banatvala J.E. Schoub B. Griffiths P.D. Mortimer P. Principles and practice of clinical virology. 6th ed. Wiley-Blackwell,
Chichester2009: 669-698
The natural host of the virus is non-human primates and the vectors in Africa are usually mosquitoes of species Aedes africanus in forest areas and Aedes aegypti in urban areas.
In humans, the acute phase occurs 3–6 days after infection and is characterized by non-specific symptoms that last up to 4 days. Fifteen percent of the patients enter a toxic phase characterized by high fever, unexplained bleeding, jaundice, and multiple organ failure, and 50% of these patients die within 2 weeks.
To confirm the disease, serological testing by way of ELISA for YF virus-specific IgM or isolation of the virus from blood samples is usually undertaken, since these are the recommended standard diagnostic tests for YF.
YF is a moving epizootic in endemic regions of tropical Africa and South America lying within a band from 15°N to 10°S of the equator and accounts for an estimated 200 000 cases of YF (with 30 000 deaths) per year globally.
After 1972, political instability in the country led to a decline in YF surveillance activities in Uganda and hence many suspected YF cases notified by the health facilities were not investigated. Following the introduction of the Integrated Disease Surveillance and Response (IDSR) strategy in Uganda in 2000,
a clinical health facility-based surveillance system for YF was instituted, though most of these reports were not accompanied by specimens to facilitate laboratory investigations. It is therefore possible that undetected human cases were occurring despite the absence of confirmed outbreaks in the country over the past four decades.
In November 2010, the Uganda Ministry of Health received reports of an outbreak of a fatal, febrile, hemorrhagic illness in northern Uganda. Multisector teams were established to support the outbreak response. This report highlights the activities and findings of the epidemiology and laboratory team during the outbreak. Entomological and reactive vaccination outcomes will be reported separately.
2. Methods
2.1 Epidemic site
A total of 15 districts in northern Uganda, including Abim, Agago, Apac, Kitgum, Kaabong, Kotido, Lamwo, Arua, Lira, Pader, Gulu, Nebbi, Napak, Dokolo, and Yumbe reported suspected cases of YF (Figure 1). The districts share borders with southern Sudan and Kenya where outbreaks have been reported in the past.
The northern region is home to a number of game reserves and parks including Kidepo Valley National Park and Murchison Falls National Park. Northern Uganda has just recovered from a nearly two-decade civil war that confined more than 90% of the population to internally displaced people's (IDP) camps. The YF outbreak occurred within 2 years after these people had returned to their original homes.
The epidemiology and laboratory team was one of the sub-committees of the national task force established to respond to the outbreak. The experts on the team included medical epidemiologists, physicians, laboratory experts, and social workers drawn from the Ministry of Health, other sectors in government, and partner organizations like the World Health Organization (WHO), Centers for Disease Control and Prevention (CDC), Médecins Sans Frontières (MSF), and the African Field Epidemiology Network (AFENET).
2.3 Epidemiology and laboratory activities
2.3.1 Epidemiology
This was a case-series investigation in which the field teams conducted detailed clinical descriptions of all suspected cases. A working case definition was developed for the initially unknown disease, and following the identification of YF using the NGS approach,
this was modified (Table 1) to facilitate the identification of additional YF cases at the health facility and community level. All new suspected YF cases were given supportive treatment
at designated treatment centers that maintained case line-lists with key variables including: identifiers, age, sex, occupation, residence, date of onset of illness, date of admission to health facility, clinical signs and symptoms, YF vaccination status, types of specimens collected, date of specimen collection, laboratory results, case classification, and case outcome.
Table 1Case definitions for the epidemiological investigation of the yellow fever outbreak in northern Uganda, 2010–2011
Period
Classification
Definition
November 8, 2010 to December 30, 2010
Suspected case
Any patient presenting with severe headache with or without fever AND at least three of the following signs or symptoms: • Gastrointestinal illness such as vomiting, hematemesis, watery/bloody diarrhea, dark stools (melena), constipation, foul smelling stools, or • Dizziness, or • General weakness, or • Convulsions, or • Any unexplained bleeding from another site
December 31, 2010 to February 9, 2011
Suspected case
Any person with acute onset of fever, with either a negative laboratory test (blood slide or RDT) for malaria or failure to respond to a full course of antimalarials AND any one of the following: • Jaundice or scleral icterus appearing within 14 days of onset of the first symptoms, or • Unexplained bleeding from either the mouth, nose, gums, skin, eyes, or stomach (gastrointestinal tract)
Probable case
Any person meeting the suspected case definition criteria WITH either: • An epidemiological link to a confirmed case or the yellow fever outbreak in northern Uganda, or • Positive post-mortem liver histopathology
Confirmed case
Any person meeting the suspected or probable case definition criteria AND one of the following: • Detection of yellow fever virus-specific IgM • Detection of yellow fever virus-specific neutralizing antibodies • Detection of yellow fever virus genome in blood or liver by PCR or NGS
RDT, rapid diagnostic test for malaria; IgM, immunoglobulin M; PCR, polymerase chain reaction; NGS, next generation sequencing.
Prior to the confirmation of YF, the types of specimens taken reflected the proposed differential diagnoses (Table 2). Hence from each suspected case, one to five blood specimens of 1–3 ml each were obtained for blood culture, viral serology and genetic testing by PCR and/or NGS, blood chemistry, full blood counts, and blood smears. Stool was also obtained in a plain container for microscopy or placed in Cary–Blair transport medium for stool culture. Liver autopsy specimens were obtained and preserved in10% formalin to facilitate YF antigen detection through immunohistochemical testing. After the confirmation of YF by NGS,
samples from new suspected cases were collected for blood chemistry, full blood counts, and YF testing by PCR and IgM, with confirmation by plaque reduction neutralization test (PRNT). Because of an ongoing hepatitis E epidemic in northern Uganda, specimens testing negative for YF were subsequently tested for hepatitis E virus and other infections, as elaborated in Table 2.
Table 2Differential diagnoses investigated during the yellow fever outbreak in northern Uganda, 2010–2011
Type of illness
Differential diagnoses
Laboratory tests undertaken
Name of Laboratory
Bacterial
Complicated shigellosis
Stool cultures
Central Public Health Laboratories, Uganda; CDC Atlanta, USA
Typhoid fever
Stool and blood cultures
Central Public Health Laboratories, Uganda; CDC Atlanta, USA
Escherichia coli (enterohemorrhagic)
Stool cultures and Shiga toxin testing
Central Public Health Laboratories, Uganda; CDC Atlanta, USA
Campylobacter jejuni
Stool cultures
Central Public Health Laboratories, Uganda; CDC Atlanta, USA
Yersinia enterocolitica
Stool cultures
Central Public Health Laboratories, Uganda; CDC Atlanta, USA
Clostridium perfringens
Stool cultures
Central Public Health Laboratories, Uganda; CDC Atlanta, USA
Anthrax
Rapid tests, blood cultures
Central Public Health Laboratories, Uganda
Plague
Rapid tests, blood cultures
Uganda Virus Research Institute Plague Laboratory, Arua; CDC Fort Collins, USA
Protozoal
Entamoeba histolytica
Stool microscopy
Central Public Health Laboratories, Uganda
Severe malaria
Blood slide for malaria parasites
Central Public Health Laboratories, Uganda
Viral
Ebola/Marburg, Lassa fever
Serology (IgM, IgG), PCR
Uganda Virus Research Institute; CDC Atlanta, USA
Dengue fever
Serology (IgM), PRNT
CDC Dengue Fever Laboratory, Puerto Rico
West Nile virus
Serology (IgM), PRNT
CDC Dengue Fever Laboratory, Puerto Rico
Yellow fever
Serology (IgM), PRNT
Uganda Virus Research Institute; CDC Atlanta, USA
Rift valley fever
Serology (IgM, IgG), PCR
Uganda Virus Research Institute; CDC Atlanta, USA
Fulminant acute viral hepatitis E
Serology (IgM), PCR
Uganda Virus Research Institute
Intoxication
Adulterated alcohol
Chemical assay (methanol)
Uganda Government Analytical Laboratory
CDC, Centers for Disease Control and Prevention; IgM, immunoglobulin M; IgG, immunoglobulin G; PCR, polymerase chain reaction; PRNT, plaque reduction neutralization test.
A Microsoft Excel data-entry screen incorporating all the variables on the YF line-list was developed and used to capture all the information on the cases reported during the outbreak. The population projections for the affected districts
Centers for Disease Control and Prevention. Epi Info database and statistics software for public health professionals conducting outbreak investigations, managing databases for public health surveillance and other tasks, and general database and statistics applications. Atlanta, GA: CDC; 2010. Available at: http://www.cdc.gov/epiinfo/ (accessed December 28, 2010).
Epidemic curves were drawn using the dates of onset to determine outbreak trends.
All epidemiological and laboratory data were collected as part of the routine outbreak investigation by national multisector teams, hence ethical clearance was not obtained.
3. Results
3.1 Outbreak investigation and response
Following prolonged outbreak investigations that lasted 40 days, laboratory confirmation of YF was made on December 18, 2010 by the CDC (Atlanta, GA, USA) using a random-primed pyrosequencing approach.
Consequently, the Ministry of Health declared an outbreak of YF in northern Uganda. A national response plan prioritizing surveillance and laboratory confirmation, case management, social mobilization and health education, and reactive vaccination was developed to control the outbreak.
3.2 Index case investigations
In Abim District, the index case was a 41-year-old male from the village of Wipolo, Aremo Parish, Morulem Sub-County. He frequented the forest to collect bamboo for sale in the local market. This forest is located near the village and most homes are located within a 0.5–1.5 km distance of the forest. His illness started on October 2, 2010 and was characterized by fever, headache, epigastric pain, hematemesis, and melena. He was not treated at any health facility and died 5 days after the onset of illness. He had not traveled out of his home village or stayed with any sick persons prior to the onset of illness. A total of five cases resulting in two deaths were reported from this family, who shared the same house. The other four affected house occupants included a 10-year-old female, a 12-year-old male who died, an eight-year-old male, and a 16-year-old male. None of the four had traveled to the forest, but they went down with a similar illness.
In Kitgum District, the index case was a 38-year-old male hunter from the village of Pudpud, Okuti Parish, Orom Sub-County. The index case village is located at the edge of a forest, and hunting gadgets were recovered from the village. His illness started on November 16, 2010 with complaints of headache, fever, epigastric pain, epistaxis, red eyes, vomiting dark blood, and passing feces with blood. He was not taken to any health facility, but was treated with unspecified medication and died on November 22, 2010. The deceased had neither traveled out of the village nor stayed with sick persons prior to the onset of illness. Three weeks after he died, the patient's mother and son fell sick with similar manifestations. This family had three cases and one death. The index case in Kitgum was not epidemiologically linked to the index case in Abim.
3.3 Epidemiological description of cases
A total of 273 suspected YF cases including 58 deaths were reported during the period October 2, 2010 to January 28, 2011 from the 15 districts in northern Uganda. The districts included Abim, Agago, Apac, Kitgum, Kaabong, Kotido, Lamwo, Arua, Lira, Pader, Gulu, Nebbi, Napak, Dokolo, and Yumbe.
However, case-based information was available for only 250 (91.6%) suspected cases and 55 (94.8%) deaths that met the initial (unknown disease) case definition. Among the cases with case-based information who met the initial case definition, 181 suspected YF cases presented with fever and unexplained bleeding or jaundice and hence met the second YF case definition; of these cases, 45 died (case fatality rate (CFR) 24.9%). The subsequent analysis therefore applies to the 181 cases and 45 deaths that met the case definition for YF.
The CFR among males (29.6%; 32/108) was nearly twice that of females (17.8%; 13/73). The cases meeting the case definition for suspected YF were reported from the 10 districts of Abim (29 cases), Agago (41 cases), Apac (one case), Gulu (three cases), Kaabong (49 cases), Kitgum (35 cases), Lamwo (14 cases), Napak (two cases), Nebbi (one case), and Pader (six cases).
Among the 181 cases that met the case definition for suspected YF, 173 (96%) underwent laboratory testing for YF, and only 13 (7.5%) were laboratory confirmed as YF; these cases came from five (50%) of the districts with cases meeting the definition for suspected YF. These districts included Agago (two cases), Abim (seven cases), Kitgum (two cases), Pader (one case), and Lamwo (one case). The CFR among YF confirmed cases was 53.8% (7/13).
The age of YF suspected cases varied from 3 months to 83 years, with a mean of 28.2 years, a standard deviation of 17.5 years, and an interquartile range of 24 years. The age among patients was almost normally distributed, with a slight skewing towards the older age groups.
3.4 Clinical presentation
Suspected YF cases presented with fever (100%), any form of unexplained bleeding (97.8%), headache (71.3%), non-bloody vomiting (59.7%), hematemesis (52.5%), epistaxis (42.0%), bloody stools (40.3%), bleeding from at least two sites (13.3%), and jaundice (11.6%).
The duration between illness onset and recovery varied from 3 to 37 days, with a median of 6 days, a mean of 9 days, and a standard deviation of 8.9 days. Amongst patients who died, the duration between onset of illness and death varied from zero to 21 days, with a median of 3 days, a mean of 4 days, and a standard deviation of 5 days.
The initial case occurred in the 39th epidemiological week of 2010 in Abim District (Figure 2). Epidemiological investigations were initiated in the 45th epidemiological week of 2010 after reports of a febrile hemorrhagic illness emerged in Abim District. In the subsequent weeks there was a gradual increase in cases reported from the 10 affected districts, reaching a plateau that stretched from the 46th to the 50th epidemiological week of 2010; the number of reported cases declined from the 51st epidemiological week of 2010 to the 6th epidemiological week of 2011. Laboratory confirmation was made in the 51st epidemiological week of 2010, a delay of 6.5 weeks after investigations commenced. Reactive vaccination was initiated in the 3rd epidemiological week of 2011.
Figure 2Suspect yellow fever cases by week of onset, northern Uganda, 2010–2011.
The overall attack rate (cases per 100 000 population) for YF in the five districts where an outbreak was confirmed was 13; this varied from 2.9 in Pader to 32.1 in Abim District (Table 3). The risk of YF transmission was highest in Morulem Sub-County in Abim District and Paimol Sub-County in Agago District (Figure 3).
Table 3Yellow fever suspected case distribution by district, northern Uganda, 2010–2011
In the five districts where YF was confirmed, the sex-specific attack rates showed that males (16.5/100 000 population) were more affected than females (9.6/10 000 population), with the risk being at least twice as high in males compared to females in the districts of Abim, Agago, and Lamwo (Table 4). In addition, the risk of YF infection increased with age (Figure 4).
Table 4Distribution of yellow fever suspect cases by sex, northern Uganda, 2010–2011
By occupation, the cases in the five districts with confirmed cases of YF included peasant farmers (36.5%), children (17.3%), housewives (13.5%), security forces (5.8%), hunters (3%), and borehole drillers (1%).
3.5 Laboratory findings
More than half of the patients had a low hemoglobin level (50.8%); 46.0% had low platelets, 41.3% had lymphocytosis, 38.1% had neutropenia, and 15.9% had leukopenia. Leukopenia with lymphocytosis was identified in 12.7% of cases (Table 5). Liver transaminases were raised in 30.2% of cases for aspartate aminotransferase (AST) and 7.9% of cases for alanine aminotransferase (ALT).
Table 5Summary of interpretations from basic investigations on suspected yellow fever cases, northern Uganda, 2010–2011
Out of 173 specimens tested, only 13 were confirmed positive for YF: four cases by PCR, eight cases by IgM and PRNT, and one case by NGS (454) testing. Molecular sequencing in one of the confirmed cases revealed 92% homology to the YF virus strain Couma (Ethiopia), belonging to the East African genotype.
Two cases from Agago District were flavivirus-indeterminate following PRNT testing. Of the samples from Kaabong District, 17.9% (5/28) were positive for hepatitis E virus-specific IgM and/or PCR. Seven patients were positive for plague using a non-validated rapid antigen test. However, all confirmatory tests for plague (i.e., blood culture, direct fluorescent antibody test, and phage lysis) were negative. One sample from a suspected case in a 62-year-old female from the neighboring district of Gulu had a positive IgM serological test for Rift Valley fever. Additional testing was undertaken on samples for the differential diagnoses listed in Table 2, but test results were negative for those diseases.
4. Discussion
During the period October 2010 to January 2011, Uganda experienced an outbreak of YF in the north, the largest ever recorded in the country. The overall attack rate in the five districts with confirmed cases was 13 cases per 100 000 population. Prior to the outbreak in northern Uganda, all of the previous five YF outbreaks in Uganda had fewer
In terms of outbreak severity, the CFR of 53.8% reported among YF confirmed cases in the northern Uganda outbreak is higher than that of Ethiopia (30%),
The CFR also measures the quality of clinical care, which is premised on a timely and definitive diagnosis. During the YF outbreak in northern Uganda, laboratory confirmation did not occur until 40 days after investigations were initiated and hence there was a delay in initiating the recommended supportive treatment
and reactive vaccination. The delay in obtaining laboratory confirmation during the YF outbreak in northern Uganda is attributable to several factors. The atypical presentation of cases was one of the reasons for the delayed confirmation, together with the observation by the physicians that patients improved on antibiotics and supportive treatment. The patients identified during the YF outbreak in northern Uganda presented with a febrile hemorrhagic illness that was characterized by fever (100%), any form of unexplained bleeding (97.8%), and headache (71.3%). Jaundice was a rare manifestation and was only reported in 11.6% of suspected YF cases. Hence, given the country's recent experience with filovirus hemorrhagic fever outbreaks,
Ebola and Marburg were the top diseases on the list of differential diagnoses (Table 2), resulting in these two diseases being prioritized for laboratory testing. The filovirus tests were, however, negative even after repeat testing was undertaken at both the Uganda Virus Research Institute (UVRI) and CDC Atlanta. Additionally, once YF was confirmed by NGS testing,
specimens had to be shipped out of the country since testing for YF was not available in Uganda, and this contributed to delayed confirmation of suspected cases. Due to the weak, case-based, laboratory-backed surveillance for YF in the country in recent years, there was no capacity for YF testing in-country when the outbreak started, hence the need to refer specimens to an international laboratory.
An investigation into the index clusters in the three districts of Abim, Agago, and Kitgum revealed that the presentation of cases was consistent with a febrile hemorrhagic illness with jaundice as a rare manifestation. However, despite the initial suspicion of a filovirus outbreak, there was inconsistent evidence of person-to-person transmission among close contacts of suspected or confirmed cases. The epidemic curve for the YF outbreak in northern Uganda showed a gradual rise in cases before reaching a plateau that lasted for close to 4 weeks, indicating a continuous common-source epidemic, a pattern that is seen in vector-borne disease outbreaks like YF due to delayed initiation of reactive vaccination. The case definition of YF that emphasizes jaundice needs to be reviewed in line with the case presentation in this outbreak to avoid missing YF cases in routine case-based surveillance.
During the outbreak, males were more affected than females and the risk of infection increased with age. In Africa, humans are seasonally exposed to YF and hence children who lack naturally acquired immunity are at high risk of disease.
Serologic testing among suspected patients during the YF outbreak in northern Uganda suggests limited natural or vaccine-induced immunity, hence adult males who undertake activities inside or close to forested areas were at higher risk of infection prior to the reactive vaccination campaign. It therefore seems likely that most human cases during the outbreak were probably infected by sylvatic vectors. Additionally, until 2 years prior to the outbreak, the civil strife in the region had confined families to IDP camps. At the time the outbreak occurred, most of the families had just returned to their original homes and hence had to clear forests and/or bush after being away for close to two decades. In view of the fact that there were at least two suspected YF cases in each of the case-series without evidence of travel to forests, this raises the possibility of urban transmission during the outbreak.
A limitation of this study arises from the fact that not all suspected YF cases underwent laboratory investigation. The few suspected cases of YF that were not tested could easily have had any of the diseases listed as differentials, particularly hepatitis E virus infection, which has been rampant in the region since 2007.
Also, 92.8% of suspected YF cases were negative for YF, indicating that the suspected case definition was very sensitive; this may be attributable to the heightened public and clinical concern regarding the unknown illness and an initially sensitive working case definition that was subsequently made specific for YF, but still included non-specific signs like epistaxis, which captured suspected cases who did not have YF. This may be a reflection of the hysteria created in the country by this outbreak. Many cases of febrile illness with any signs of bleeding were reported as suspected YF cases. Most of these were not confirmed as YF cases and were not investigated further. The high CFR in non-YF cases deserves further investigations.
Following the re-emergence of YF in Uganda, we recommend that case-based laboratory-backed surveillance for YF is strengthened. YF should be considered as a differential in all cases presenting with a febrile hemorrhagic illness, even in the absence of jaundice. In addition, the in-country capacity for YF testing should be re-established. A nationwide risk assessment should be undertaken to inform the national YF control strategy.
Acknowledgements
We would like to recognize the contribution of the following individuals and organizations to the investigation and response to the outbreak: the district task force committees for Kitgum, Agago, Lamwo, Pader, Abim, and Kaabong districts; all members of the national task force including the staff from the Ministry of Health Headquarters in Uganda, Central Public Health Laboratories, National Medical Stores, Uganda National Expanded Program on Immunization (UNEPI), and Uganda Virus Research Institute (UVRI), including the Plague Laboratory in Arua; the staff from other government sectors including: Makerere University School of Public Health, Makerere University Faculty of Medicine, Ministry of Agriculture Animal Industry and Fisheries, and the Uganda Government Chemists; WHO staff from the Uganda Country Office, AFRO, IST-East and Southern Africa, and HQ-Geneva; CDC staff in Entebbe, Atlanta, and Fort Collins, including Christopher Taylor, Aimee Geissler, Sundeep Gupta, Jeff Borchert, Mary Crabtree, Marc Fischer, Kevin Griffith, Nicole Lindsey, Paul Mead, Barry Miller, Erin Staples, John Besser, Cheryl Bopp, Collette Fitzgerald, Peter Gerner-Smidt, Gerardo Gomez, Michele Parsons, James Prucker, Janet Pruckler, Sherricka Simington, Marty Schriefer, Deborah Talkington, Anne Whitney, Laura McMullan, Shelley Campbell, Gregory Kocher, Zachary Reed, Tara Sealy, Adam MacNeil, Pierre Rollin, and Stuart Nichol; staff from other partner organizations including UNICEF, AFENET, MSF-Spain, Jonathan Polonsky from MSF-Holland, Uganda Red Cross, World Vision, RESPOND, PREDICT, and Conservation Through Public Health.
Role of the funding source: The outbreak response was funded by the National Task Force for Epidemic Control. The funds were principally drawn from the Government of Uganda budget with support from the health development partners. The corresponding author had full access to the investigation data and had the final responsibility for the decision to submit for publication.
Conflicts of interest: There were no potential conflicts of interest tendered by any of the authors.
References
Schoub B.D.
Venter M.
Flaviviruses.
in: Zuckerman A.J. Banatvala J.E. Schoub B. Griffiths P.D. Mortimer P. Principles and practice of clinical virology. 6th ed. Wiley-Blackwell,
Chichester2009: 669-698
Centers for Disease Control and Prevention. Epi Info database and statistics software for public health professionals conducting outbreak investigations, managing databases for public health surveillance and other tasks, and general database and statistics applications. Atlanta, GA: CDC; 2010. Available at: http://www.cdc.gov/epiinfo/ (accessed December 28, 2010).
00116-6&id=gr1.jpg){kind=link}
00116-6&id=gr2.jpg){kind=link}
00116-6&id=gr3.jpg){kind=link}
00116-6&id=gr4.jpg){kind=link}