- Email: [email protected]
Under-ascertainment, under-reporting and timeliness of Iranian communicable disease surveillance system for zoonotic diseases
p u b l i c h e a l t h 1 5 4 ( 2 0 1 8 ) 1 3 0 e1 3 5
Available online at www.sciencedirect.com
Public Health journal homepage: www.elsevier.com/puhe
Original Research
Under-ascertainment, under-reporting and timeliness of Iranian communicable disease surveillance system for zoonotic diseases P.A. Kazerooni a, M. Fararouei a,*, M. Nejat b, M. Akbarpoor c, Z. Sedaghat d a
HIV/AIDs Research Center, Shiraz University of Medical Sciences, Shiraz, Iran Student Research Center for Health Sciences, Department of Epidemiology, School of Health, Shiraz University of Medical Sciences, Shiraz, Iran c Health Affairs, Shiraz University of Medical Sciences, Shiraz, Iran d School of Health, Shiraz University of Medical Sciences, Shiraz, Iran b
article info
abstract
Article history:
Objectives: The important steps in controlling infectious diseases are fast detection, proper
Received 5 May 2017
treatment and on-time reporting of cases to the appropriate authorities. This study was
Received in revised form
conducted to evaluate the quality of Iranian communicable diseases surveillance system
26 October 2017
(CDSS) for zoonotic diseases in rural areas of Fars province (Iran's third largest province).
Accepted 30 October 2017
Study design: The three most important evaluation indices of CDSS, namely underascertainment, under-reporting and timeliness, for the three most common zoonotic diseases were measured using independent data source obtained from door-to-door in-
Keywords:
terviews and patients' medical records.
Zoonotic diseases
Methods: Interviews were conducted with 48,771 households in rural areas of Fars province
Surveillance
from April 2014 to March 2015. The medical and registration data were obtained from the
Sensitivity
CDSS and medical or health centres.
Timeliness
Results: Under-ascertainment, total under-reporting and timeliness (delay from the time of onset of symptoms to the time of visiting a medical or health centre, to the time of reporting visited cases to the highest level of CDSS) for leishmaniasis were 19.6%, 42.5% and 81.61 days (48.95 due to system delay), respectively. The corresponding indices for brucellosis were 0%, 41.8% and 56.5 days (22 due to system delay), respectively. For animal bite, the corresponding indices were 7.83%, 13.07% and less than 72 h, respectively (no system delay). Conclusions: Although the status of case reporting and timeliness of surveillance system in public sectors providing medical services are clearly better than those of the private sectors, the indices are far from the level needed by CDSS to be able to detect and handle epidemics on time. Training health personnel, especially physicians, from public and private sectors to secure their cooperation along with routine and indebt evaluation are necessary to improve CDSS in Iran. © 2017 The Royal Society for Public Health. Published by Elsevier Ltd. All rights reserved.
* Corresponding author. Shiraz University of Medical Sciences, Zand St., Shiraz, Iran. E-mail address: [email protected] (M. Fararouei). https://doi.org/10.1016/j.puhe.2017.10.029 0033-3506/© 2017 The Royal Society for Public Health. Published by Elsevier Ltd. All rights reserved.
p u b l i c h e a l t h 1 5 4 ( 2 0 1 8 ) 1 3 0 e1 3 5
Introduction According to WHO, ‘health surveillance is an ongoing systematic collection, analysis, interpretation and dissemination of health-related data to plan, implement and evaluate public health programs and interventions’.1 According to the ‘International Health Regulations’ published by WHO, it is necessary for all member countries to periodically assess and improve their own surveillance systems.2,3 Two essential steps in surveillance of diseases at a local or national level are complete and fast detection and reporting of incidents of communicable diseases. To achieve this, setting a wellorganized and reliable national diseases surveillance system (DSS) is crucial.4 As the result, the quality of any surveillance system is defined by several indices including timeliness and sensitivity of the reporting procedures.2 Sensitivity of a DSS is defined as ‘the proportion of cases of a disease (or other health-related event) detected by the surveillance system’.1 Timeliness is defined as the mean of elapsing time between different steps in DSS.2 Several published studies on the quality of Iranian DSS have shown serious challenges in the sensitivity and timeliness of the system.5e7 As a result, conducting evaluation studies on surveillance systems based on the indices mentioned above are essential to ensure that the DSS is performing effectively.1 According to the centre for communicable disease control (CDC) based in the Iranian ministry of health, communicable diseases are categorized in two groups: diseases which require immediate reporting (within 24 h) to the highest level of communicable DSS (CDSS) and those which require periodic reporting within up to a month to the highest level of CDSS. Accordingly, public and private sectors are to register cases and report the notifiable diseases to the county's CDSS office and then to the provincial health centres via phone call and specially designed forms. In the next step, provincial health centres merge and analyse the data and send the results to the ministerial office of CDC. The procedures of CDSS which are similar in all provinces in Iran are fully explained elsewhere.6 This is an evaluation study conducted to evaluate the quality of Iranian CDSS in rural areas of a large province located at the southern part of the country.
Methods This is a population-based evaluation study to measure the three main indices of a surveillance system (under-ascertainment, under-reporting and timeliness) for syndromic zoonotic diseases, namely leishmaniasis and brucellosis. In addition, the indices were measured for the surveillance of animal bites (a significant index for rabies control). The required information about the incidence of the above diseases in rural areas was gathered independent to the CDSS via door-to-door interviews using an interview-administered questionnaire. The sampling was done based on a population registry database available at the deputy of heath office in Shiraz University of Medical Sciences, Shiraz, Iran. Interviews were conducted at house doors by a team of trained public
131
health nurses. Two health experts, who had official access to all medical records at the health and medical centres, collected the required data via direct visits to the centres. A year prior to the current research, a pilot study with relatively large sample size was conducted on the same subject to evaluate and revise the questionnaires and the study procedures.6 Owing to time-dependency of the occurrence of the selected diseases and the difference in timing of the sampling of the pilot and the main study, the data of the pilot study were not included in the analysis of the present study and like other parts of the province, the rural population of Marvdasht county (the county in which the pilot study was conducted) was also sampled for the present study. In addition, Marvdasht is located next to capital of Fars province (Shiraz), provided with a network of relatively well-equipped and wellstaffed health facilities. With regard to the above issues, the results of pilot and the present study were considered incomparable.
Definitions Different terminologies are provided for indices of DSS.1,2,8 Throughout the text, the terms and definitions used are as follow: as the result of having no information on asymptomatic patients, all indices are calculated based on the symptomatic cases in the population. As one of the main indicators of sensitivity, under-ascertainment is calculated as the proportion of symptomatic cases who did not visit any health/ medical care provider because of their defined symptoms. Under-reporting is the other important index for sensitivity which is calculated as the proportion of symptomatic cases who had visited at least one health/medical service provider due to their defined symptoms but were not reported to the higher level of surveillance system.9 Multiplication factor (MF) was also calculated to estimate the number of incident symptomatic cases during the defined study period in the study population.9
Case definition The symptoms were defined based on the latest Iranian ministry of health's guidelines for reporting communicable diseases, which are inspired by practicality and feasibility. The definitions for compatible cases of leishmaniasis, brucellosis and animal bite are as follows: Leishmaniasis: Clinically compatible case: A person having papules or cutaneous ulcer with no apparent external cause. These cases are to be reported monthly to the provincial CDC office. The cases are to be confirmed by positive smear for leishman bodies. Brucellosis: Clinically compatible case: A person with classic symptoms and positive serologic test is to be reported monthly to the provincial CDC office. Though not very accurate compared to PCR, the final diagnosis of brucellosis is to be made by serologic tests. This is due to limited financial and geographical accessibility of the population to advanced medical laboratory services. Rabies: A person having animal bite. The bitten individuals are to be reported monthly to the provincial CDC office and to
132
p u b l i c h e a l t h 1 5 4 ( 2 0 1 8 ) 1 3 0 e1 3 5
provincial health centre and ministry of health during the study period. The indices were calculated as following: underascertainment was calculated as the ration of those with symptoms who did not visit any health or medical care provider during the course of their condition to all symptomatic patients. Under-reporting was calculated as the ratio of cases who visited a health provider but were not reported to each level of surveillance system to the total number of cases who visited (at least once) a healthcare provider. Assuming that the reported cases are a representative sample of all symptomatic cases in the population, multiplication factor (a factor to be used to calculate a relatively valid estimate of the total symptomatic cases in the population regardless of visiting a health provider) is also calculated. Timeliness was calculated as the summation of the elapsing time between onset of symptoms to the first presentation of the patients to a medical care provider and from the first presentation of the patients to the date in which the case was reported to the provincial health centre.
the highest level at the ministry of health. All cases are to receive immediate preventive measures, mainly vaccination, by the vaccination centres. The only interested factor for evaluating surveillance of rabies was the number of individuals who were bitten by an animal. It should be noted that no attempt was made to get access to the final diagnosis of the patients due to ethical considerations.
Settings The rural population of Fars, the third Iranian largest province, was selected as the source population. The province is located at the southern part of the country. Houses were selected randomly, based on which 500 houses (48,771 individuals) were selected. An interview-administered questionnaire was used to obtain data on the households' health status from mothers or other adult women by a trained female nurse. Accordingly, the mothers were asked if there were any of the interested symptoms among their family members.
The population-based data collection
Results
The interview was aimed to find out the defined symptoms of the selected diseases (leishmaniasis, brucellosis and animal bite) which occurred during April 2014 to March 2015 (the past 12 months to the date of interview). If a household member suffered from any of the defined sign or symptoms, the second interview would be conducted with the affected member of the family. The required information about the patients (type of the symptoms, name and address of health or medical centres that they presented themselves and exact date of their presentation) was obtained from both patient's self-reported data and medical records at the centres. The interviews were carried out by 50 groups, each consisted of two experienced public health nurses, one man and one woman. The aims of the study and the confidentiality of the patient's information were explained to the participants before interview. As a significant number of the participants were illiterate, a verbal informed consent was obtained from responding women and any affected family member with the history of interested symptoms just before the interview was started.
In total, 48,771 individuals were selected as the study sample. The numbers of syndromic cases of leishmaniasis, brucellosis and animal bites were 199, 55 and 166, respectively. The baseline characteristics of participants with and without the interested symptoms are shown in Table 1.
Under-ascertainment and under-reporting of CDSS Leishmaniasis: Out of 199 suspected cases of leishmaniasis, who were detected during door-to-door interviews, 160 cases (80.40%) presented themselves to at least one medical care provider to receive diagnosis and treatment services. Of these, most patients visited public rural health centres or rural health houses (65.69%). Out of patients who visited a health or medical centre, 92 (57.5%) cases were registered and reported to the highest level of surveillance system. Accordingly, the rate of total under-reporting and MF of CDSS for leishmaniasis was 42.5% and 2.16 respectively (Table 2). Brucellosis: All 55 syndromic cases of brucellosis (100%) presented themselves to at least a healthcare provider to receive diagnosis and treatment services. However, only 32 cases (58.18%) were reported to the highest level of surveillance system. Accordingly, the rate of total under-reporting and MF of CDSS for brucellosis were 41.82% and 1.72 respectively (Table 2). Animal bites: Of 166 participants who reported animal bites, 153 cases (92.17%) presented themselves to at least one medical care provider to receive treatment services. Again, most patients (88.16%) visited rural health centres and rural
The health systemebased information Data on the registered cases with the interested symptoms were obtained from rural and urban clinics and health centres. To calculate under-ascertainment, under-reporting and timeliness of CDSS for the interested diseases, required information was obtained from patients and health records, providing the patients were registered in the medical centres or were reported to the CDC offices in the headquarters of
Table 1 e Baseline characteristics of detected cases in rural areas of Fars province. Disease
Leishmaniasis Brucellosis Animal bites
Number of households
Number of detected cases
Sex ratio (M/F)
48,771 48,771 48,771
199 55 166
0.91 1.20 2.39
Age (years) Median
SD
Mean
20.75 39.78 31.96
19.73 17.80 20.70
12.00 35.00 28.00
Incidence rate (Per 100,000) 408.03 112.77 340.37
Table 2 e Sensitivity of Iranian communicable diseases surveillance system (CDSS) according to different levels of surveillance system for leishmaniasis, brucellosis and animal bites (rabies). Disease
Cases visited Number Number of detected health service of visits symptomatic provider [B] cases [A] (underascertainment, 1-B/A)
Leishmaniasis
199
160 80.40% (19.60%)
274
Service provider
Laboratory Brucellosis
55
55 100% (0.00%)
123
Rural health centre/ Health houses General practitioner in private sector Specialist practitioner in private sector Hospital Laboratory
Disease
Suspected cases of Rabies
180 65.69% 7 2.55% 29 10.58% 25 9.12% 33 12.04% 30 24.39% 15 12.19% 14 11.38% 9 7.32% 55 44.71
92
48.89%
0
0
0
0
0
0
1
96.97%
20
33.33%
1
93.33%
3
78.57%
2
77.78%
8
85.45%
92
92
0%
42.5% (2.16)
32
32
0%
41.82% (1.72)
Number of detected symptomatic cases [A]
Cases visited health service provider [B] (underascertainment, 1-B/A)
Number of visits
Service provider
Number of visits [C]
Reported to rabies vaccination centre [D]
Under-reporting to a rabies vaccination centre [1-D/C]
Reported to a rabies vaccination centre (duplicates excluded) [E]
Total under-reporting [1-E/B] (MF¼A/E)
166
153 92.17% (7.83%)
169
Rural health centre/ Health houses General practitioner in private sector Specialist practitioner in private sector Hospital
149 88.16% 0
130
12.75%
133
13.07% (1.25)
20 11.83%
e
0
e
16
20%
133
CDSS, communicable diseases surveillance system; MF, multiplication factor.
0
0
p u b l i c h e a l t h 1 5 4 ( 2 0 1 8 ) 1 3 0 e1 3 5
Rural health centre/ Health houses General practitioner in private sector Specialist practitioner in private sector Hospital
Number Reported to Under-reporting Reported cases Reported to Under Total of visits county's to the next to county's health the highest -reporting under-reporting [C] health level [1-D/C] centre (duplicates level [F] to the [1-F/B] (MF¼A/F) centre [D] excluded) [E] highest level [1-F/E]
134
p u b l i c h e a l t h 1 5 4 ( 2 0 1 8 ) 1 3 0 e1 3 5
Table 3 e Timeliness of Iranian CDSS for leishmaniasis and brucellosis. Number of participants
Number of detected cases
Cases reported to the highest level
Time between the onset of symptoms to the first presentation at a medical care centre (days) [Mean (med)]
Time between the first presentation at a medical care centre to being reported to the highest level of CDSS (days) [Mean (med)]
Total elapsing time (days) [Mean (med)]
Leishmaniasis
48,771
199
92
32.66 (29)
48.95 (30)
81.61 (63)
Brucellosis
48,771
55
32
34.5 (24)
22.0 (20.5)
56.5 (48)
Disease
Abbreviation: CDSS, communicable diseases surveillance system; med, median.
medical or healthcare provider. Only about 57% of the leishmaniasis cases who sought medical services were registered and reported to the highest level of CDSS. The average elapsing time for reporting leishmaniasis cases was about 80 days, which is too long to make CDSS able to implement preventive measures effectively. The results of the present study also suggested that all suspected patients of brucellosis sought diagnosis and treatment services and presented themselves to at least a medical centre. On the other hand, only 54.55% of the brucellosis cases who presented themselves to health or medical centres were reported to the higher levels. The elapsing time for brucellosis cases who sought health and medical services and were registered and reported to the highest level of surveillance system was too long (about 2 months) to make CDSS able to initiate control programs. Of all individuals who were bitten by an animal, 92.17% presented themselves to a medical centre. According to the results, 86.93% of the cases of animal bites were reported to the highest level of surveillance system and predominantly were vaccinated within less than 72 h since the animal bite. In spite of the relatively good sensitivity and timeliness of reporting animal bites, because of the fatal importance of ontime detection and vaccination of the cases, further improvement in quality of surveillance system is necessary. Regarding the high incidence rate of leishmaniasis, brucellosis and animal bites in Fars province, it seems that improvement in the timeliness and sensitivity of reporting cases is essential for effective controlling programs. It worth noticing that only a few studies have been conducted on the quality of the Iranian CDSS.5e7 The results of similar studies in the United States (2008, 2011) has shown that sensitivity of case reporting for brucellosis and rabies were 25% higher and 7% lower than the corresponding results of the present study respectively.8,10 Based on the results of another study in Greece (2010), sensitivity for
health houses. Of 153 cases who presented themselves to a health provider, 133 (86.93%) cases were registered and reported to the highest level of surveillance system. Accordingly, the rate of total under-reporting and MF of animal bites for CDSS was 13.07% and 1.25 respectively (Table 2).
Timeliness of surveillance system Leishmaniasis: The average elapsed time from the onset of symptoms to the first patient's presentation to a medical care centre was 32.66 days. The average elapsed time from patients' first presentation at the medical centres to the time of reporting cases to the highest level of surveillance system was 48.95 days. Accordingly, the total elapsing time between the onset of symptoms and reporting cases to the highest level of the health care system was on average 81.61 days (Table 3). Brucellosis: The average elapsed time between the onset of symptoms of brucellosis and the first patients' presentation to a medical care centre was 34.5 days. The average elapsed time between patients' first presentation to medical care centres and reporting cases to the highest level of surveillance system was 22 days. Accordingly, the average total elapsing time between the onset of symptoms and reporting cases to the highest level of the surveillance system was 56.5 days (Table 3). Animal bites: Only 84.96% of the detected cases were vaccinated within 24 h since animal bite. About 9.02% of the cases were vaccinated within 24e72 h and 6.02% received vaccination after 72 h from the date of animal bite (Table 4).
Discussion The results of this evaluation study suggested that more than 19% of symptomatic leishmaniasis cases did not visit a
Table 4 e Timeliness of Iranian CDSS for animal bites (rabies). Number of participants 48,771
Number of detected cases
Number referred to rabies vaccination centres
166
133
Abbreviation: CDSS, communicable diseases surveillance system.
The elapsing time between animal bite and vaccination [n (%)] <24 h 113 (84.96)%
24e72 h
>72 h
12 (9.02%)
8 (6.02%)
p u b l i c h e a l t h 1 5 4 ( 2 0 1 8 ) 1 3 0 e1 3 5
both brucellosis and leishmaniasis reported at about 74%, which is significantly higher than the corresponding rate obtained for Iranian CDSS.11 Strength and weakness: The present study was conducted in rural areas of a large province in the southern part of Iran (Fars province). To calculate the study indices, a populationbased door-to-door case finding was conducted to obtain required information independent to the CDSS database. No clinical information was obtained from medical and health centres that the patients visited during their course of disease. As the result, only symptomatic cases were detected and no attempt was made to confirm the status of diseases among the participants.
Conclusion Based on the results of the present study, the quality of case reporting for notifiable zoonotic diseases is relatively low. Disease surveillance indices must be routinely measured and improved at both governmental and private sectors. Training health personnel, especially physicians, about the objectives of the surveillance system and its importance on a periodic base is necessary. Routine evaluation of the performance of CDSS can effectively improve the validity and efficiency of the system.
Author statements Ethical approval This study was approved by the Shiraz University of Medical Sciences ethical review board.
Funding The present study was financially supported by Shiraz University of Medical Sciences, Shiraz, Iran.
135
Competing interests None declared.
references
1. German RR. Updated guidelines for evaluating public health surveillance systems. MMWR Recomm Rep 2001;50(1e35). 2. Organization, W.H. Communicable disease surveillance and response systems: guide to monitoring and evaluating. 2006. 3. Organization, W.H. International health regulations (2005). World Health Organization; 2008. 4. Heymann DL. Control of communicable diseases manual. American Public Health Association; 2008. 5. Fararouei M, et al. National guidelines for outbreak investigation: an evaluation study/Lignes directrices re d'investigation des flambe es: une e tude nationales en matie valuation. East Mediterr Health J 2013;19(9):816. d'e 6. Nejat M, et al. Surveillance of zoonotic diseases in Marvdasht city, Fars province. J Health Sci Surveill Syst 2016;4(2):95e102. 7. Fararouei M, Marzban M, Shahraki G. Completeness of cancer registry data in a small Iranian province a captureerecapture approach. Health Inf Manag J 2016. 1833358316668605. 8. Sickbert-Bennett EE. Completeness of communicable disease reporting, North Carolina, USA, 1995e1997 and 2000e2006Volume 17, number 1dJanuary 2011. Emerg Infect Dis J 2011. 9. Gibbons CL, Mangen M-JJ, Plass D, Havelaar AH, Brooke RJ, Kramarz P, et al. Measuring underreporting and underascertainment in infectious disease datasets: a comparison of methods. BMC Public Health 2014;14(1):147. 10. Thiede H, Close NS, Koepsell J, Baer A, Duchin JS. Completeness of reporting of rabies postexposure prophylaxis in King County, Washington. J Public Health Manag Pract 2008;14(5):448e53. 11. Jelastopulu E, Merekoulias G, Alexopoulos E. Underreporting of communicable diseases in the prefecture of Achaia, western Greece, 1999e2004-missed opportunities for early intervention. Euro Surveill 2010;15(21):19579.
Available online at www.sciencedirect.com
Public Health journal homepage: www.elsevier.com/puhe
Original Research
Under-ascertainment, under-reporting and timeliness of Iranian communicable disease surveillance system for zoonotic diseases P.A. Kazerooni a, M. Fararouei a,*, M. Nejat b, M. Akbarpoor c, Z. Sedaghat d a
HIV/AIDs Research Center, Shiraz University of Medical Sciences, Shiraz, Iran Student Research Center for Health Sciences, Department of Epidemiology, School of Health, Shiraz University of Medical Sciences, Shiraz, Iran c Health Affairs, Shiraz University of Medical Sciences, Shiraz, Iran d School of Health, Shiraz University of Medical Sciences, Shiraz, Iran b
article info
abstract
Article history:
Objectives: The important steps in controlling infectious diseases are fast detection, proper
Received 5 May 2017
treatment and on-time reporting of cases to the appropriate authorities. This study was
Received in revised form
conducted to evaluate the quality of Iranian communicable diseases surveillance system
26 October 2017
(CDSS) for zoonotic diseases in rural areas of Fars province (Iran's third largest province).
Accepted 30 October 2017
Study design: The three most important evaluation indices of CDSS, namely underascertainment, under-reporting and timeliness, for the three most common zoonotic diseases were measured using independent data source obtained from door-to-door in-
Keywords:
terviews and patients' medical records.
Zoonotic diseases
Methods: Interviews were conducted with 48,771 households in rural areas of Fars province
Surveillance
from April 2014 to March 2015. The medical and registration data were obtained from the
Sensitivity
CDSS and medical or health centres.
Timeliness
Results: Under-ascertainment, total under-reporting and timeliness (delay from the time of onset of symptoms to the time of visiting a medical or health centre, to the time of reporting visited cases to the highest level of CDSS) for leishmaniasis were 19.6%, 42.5% and 81.61 days (48.95 due to system delay), respectively. The corresponding indices for brucellosis were 0%, 41.8% and 56.5 days (22 due to system delay), respectively. For animal bite, the corresponding indices were 7.83%, 13.07% and less than 72 h, respectively (no system delay). Conclusions: Although the status of case reporting and timeliness of surveillance system in public sectors providing medical services are clearly better than those of the private sectors, the indices are far from the level needed by CDSS to be able to detect and handle epidemics on time. Training health personnel, especially physicians, from public and private sectors to secure their cooperation along with routine and indebt evaluation are necessary to improve CDSS in Iran. © 2017 The Royal Society for Public Health. Published by Elsevier Ltd. All rights reserved.
* Corresponding author. Shiraz University of Medical Sciences, Zand St., Shiraz, Iran. E-mail address: [email protected] (M. Fararouei). https://doi.org/10.1016/j.puhe.2017.10.029 0033-3506/© 2017 The Royal Society for Public Health. Published by Elsevier Ltd. All rights reserved.
p u b l i c h e a l t h 1 5 4 ( 2 0 1 8 ) 1 3 0 e1 3 5
Introduction According to WHO, ‘health surveillance is an ongoing systematic collection, analysis, interpretation and dissemination of health-related data to plan, implement and evaluate public health programs and interventions’.1 According to the ‘International Health Regulations’ published by WHO, it is necessary for all member countries to periodically assess and improve their own surveillance systems.2,3 Two essential steps in surveillance of diseases at a local or national level are complete and fast detection and reporting of incidents of communicable diseases. To achieve this, setting a wellorganized and reliable national diseases surveillance system (DSS) is crucial.4 As the result, the quality of any surveillance system is defined by several indices including timeliness and sensitivity of the reporting procedures.2 Sensitivity of a DSS is defined as ‘the proportion of cases of a disease (or other health-related event) detected by the surveillance system’.1 Timeliness is defined as the mean of elapsing time between different steps in DSS.2 Several published studies on the quality of Iranian DSS have shown serious challenges in the sensitivity and timeliness of the system.5e7 As a result, conducting evaluation studies on surveillance systems based on the indices mentioned above are essential to ensure that the DSS is performing effectively.1 According to the centre for communicable disease control (CDC) based in the Iranian ministry of health, communicable diseases are categorized in two groups: diseases which require immediate reporting (within 24 h) to the highest level of communicable DSS (CDSS) and those which require periodic reporting within up to a month to the highest level of CDSS. Accordingly, public and private sectors are to register cases and report the notifiable diseases to the county's CDSS office and then to the provincial health centres via phone call and specially designed forms. In the next step, provincial health centres merge and analyse the data and send the results to the ministerial office of CDC. The procedures of CDSS which are similar in all provinces in Iran are fully explained elsewhere.6 This is an evaluation study conducted to evaluate the quality of Iranian CDSS in rural areas of a large province located at the southern part of the country.
Methods This is a population-based evaluation study to measure the three main indices of a surveillance system (under-ascertainment, under-reporting and timeliness) for syndromic zoonotic diseases, namely leishmaniasis and brucellosis. In addition, the indices were measured for the surveillance of animal bites (a significant index for rabies control). The required information about the incidence of the above diseases in rural areas was gathered independent to the CDSS via door-to-door interviews using an interview-administered questionnaire. The sampling was done based on a population registry database available at the deputy of heath office in Shiraz University of Medical Sciences, Shiraz, Iran. Interviews were conducted at house doors by a team of trained public
131
health nurses. Two health experts, who had official access to all medical records at the health and medical centres, collected the required data via direct visits to the centres. A year prior to the current research, a pilot study with relatively large sample size was conducted on the same subject to evaluate and revise the questionnaires and the study procedures.6 Owing to time-dependency of the occurrence of the selected diseases and the difference in timing of the sampling of the pilot and the main study, the data of the pilot study were not included in the analysis of the present study and like other parts of the province, the rural population of Marvdasht county (the county in which the pilot study was conducted) was also sampled for the present study. In addition, Marvdasht is located next to capital of Fars province (Shiraz), provided with a network of relatively well-equipped and wellstaffed health facilities. With regard to the above issues, the results of pilot and the present study were considered incomparable.
Definitions Different terminologies are provided for indices of DSS.1,2,8 Throughout the text, the terms and definitions used are as follow: as the result of having no information on asymptomatic patients, all indices are calculated based on the symptomatic cases in the population. As one of the main indicators of sensitivity, under-ascertainment is calculated as the proportion of symptomatic cases who did not visit any health/ medical care provider because of their defined symptoms. Under-reporting is the other important index for sensitivity which is calculated as the proportion of symptomatic cases who had visited at least one health/medical service provider due to their defined symptoms but were not reported to the higher level of surveillance system.9 Multiplication factor (MF) was also calculated to estimate the number of incident symptomatic cases during the defined study period in the study population.9
Case definition The symptoms were defined based on the latest Iranian ministry of health's guidelines for reporting communicable diseases, which are inspired by practicality and feasibility. The definitions for compatible cases of leishmaniasis, brucellosis and animal bite are as follows: Leishmaniasis: Clinically compatible case: A person having papules or cutaneous ulcer with no apparent external cause. These cases are to be reported monthly to the provincial CDC office. The cases are to be confirmed by positive smear for leishman bodies. Brucellosis: Clinically compatible case: A person with classic symptoms and positive serologic test is to be reported monthly to the provincial CDC office. Though not very accurate compared to PCR, the final diagnosis of brucellosis is to be made by serologic tests. This is due to limited financial and geographical accessibility of the population to advanced medical laboratory services. Rabies: A person having animal bite. The bitten individuals are to be reported monthly to the provincial CDC office and to
132
p u b l i c h e a l t h 1 5 4 ( 2 0 1 8 ) 1 3 0 e1 3 5
provincial health centre and ministry of health during the study period. The indices were calculated as following: underascertainment was calculated as the ration of those with symptoms who did not visit any health or medical care provider during the course of their condition to all symptomatic patients. Under-reporting was calculated as the ratio of cases who visited a health provider but were not reported to each level of surveillance system to the total number of cases who visited (at least once) a healthcare provider. Assuming that the reported cases are a representative sample of all symptomatic cases in the population, multiplication factor (a factor to be used to calculate a relatively valid estimate of the total symptomatic cases in the population regardless of visiting a health provider) is also calculated. Timeliness was calculated as the summation of the elapsing time between onset of symptoms to the first presentation of the patients to a medical care provider and from the first presentation of the patients to the date in which the case was reported to the provincial health centre.
the highest level at the ministry of health. All cases are to receive immediate preventive measures, mainly vaccination, by the vaccination centres. The only interested factor for evaluating surveillance of rabies was the number of individuals who were bitten by an animal. It should be noted that no attempt was made to get access to the final diagnosis of the patients due to ethical considerations.
Settings The rural population of Fars, the third Iranian largest province, was selected as the source population. The province is located at the southern part of the country. Houses were selected randomly, based on which 500 houses (48,771 individuals) were selected. An interview-administered questionnaire was used to obtain data on the households' health status from mothers or other adult women by a trained female nurse. Accordingly, the mothers were asked if there were any of the interested symptoms among their family members.
The population-based data collection
Results
The interview was aimed to find out the defined symptoms of the selected diseases (leishmaniasis, brucellosis and animal bite) which occurred during April 2014 to March 2015 (the past 12 months to the date of interview). If a household member suffered from any of the defined sign or symptoms, the second interview would be conducted with the affected member of the family. The required information about the patients (type of the symptoms, name and address of health or medical centres that they presented themselves and exact date of their presentation) was obtained from both patient's self-reported data and medical records at the centres. The interviews were carried out by 50 groups, each consisted of two experienced public health nurses, one man and one woman. The aims of the study and the confidentiality of the patient's information were explained to the participants before interview. As a significant number of the participants were illiterate, a verbal informed consent was obtained from responding women and any affected family member with the history of interested symptoms just before the interview was started.
In total, 48,771 individuals were selected as the study sample. The numbers of syndromic cases of leishmaniasis, brucellosis and animal bites were 199, 55 and 166, respectively. The baseline characteristics of participants with and without the interested symptoms are shown in Table 1.
Under-ascertainment and under-reporting of CDSS Leishmaniasis: Out of 199 suspected cases of leishmaniasis, who were detected during door-to-door interviews, 160 cases (80.40%) presented themselves to at least one medical care provider to receive diagnosis and treatment services. Of these, most patients visited public rural health centres or rural health houses (65.69%). Out of patients who visited a health or medical centre, 92 (57.5%) cases were registered and reported to the highest level of surveillance system. Accordingly, the rate of total under-reporting and MF of CDSS for leishmaniasis was 42.5% and 2.16 respectively (Table 2). Brucellosis: All 55 syndromic cases of brucellosis (100%) presented themselves to at least a healthcare provider to receive diagnosis and treatment services. However, only 32 cases (58.18%) were reported to the highest level of surveillance system. Accordingly, the rate of total under-reporting and MF of CDSS for brucellosis were 41.82% and 1.72 respectively (Table 2). Animal bites: Of 166 participants who reported animal bites, 153 cases (92.17%) presented themselves to at least one medical care provider to receive treatment services. Again, most patients (88.16%) visited rural health centres and rural
The health systemebased information Data on the registered cases with the interested symptoms were obtained from rural and urban clinics and health centres. To calculate under-ascertainment, under-reporting and timeliness of CDSS for the interested diseases, required information was obtained from patients and health records, providing the patients were registered in the medical centres or were reported to the CDC offices in the headquarters of
Table 1 e Baseline characteristics of detected cases in rural areas of Fars province. Disease
Leishmaniasis Brucellosis Animal bites
Number of households
Number of detected cases
Sex ratio (M/F)
48,771 48,771 48,771
199 55 166
0.91 1.20 2.39
Age (years) Median
SD
Mean
20.75 39.78 31.96
19.73 17.80 20.70
12.00 35.00 28.00
Incidence rate (Per 100,000) 408.03 112.77 340.37
Table 2 e Sensitivity of Iranian communicable diseases surveillance system (CDSS) according to different levels of surveillance system for leishmaniasis, brucellosis and animal bites (rabies). Disease
Cases visited Number Number of detected health service of visits symptomatic provider [B] cases [A] (underascertainment, 1-B/A)
Leishmaniasis
199
160 80.40% (19.60%)
274
Service provider
Laboratory Brucellosis
55
55 100% (0.00%)
123
Rural health centre/ Health houses General practitioner in private sector Specialist practitioner in private sector Hospital Laboratory
Disease
Suspected cases of Rabies
180 65.69% 7 2.55% 29 10.58% 25 9.12% 33 12.04% 30 24.39% 15 12.19% 14 11.38% 9 7.32% 55 44.71
92
48.89%
0
0
0
0
0
0
1
96.97%
20
33.33%
1
93.33%
3
78.57%
2
77.78%
8
85.45%
92
92
0%
42.5% (2.16)
32
32
0%
41.82% (1.72)
Number of detected symptomatic cases [A]
Cases visited health service provider [B] (underascertainment, 1-B/A)
Number of visits
Service provider
Number of visits [C]
Reported to rabies vaccination centre [D]
Under-reporting to a rabies vaccination centre [1-D/C]
Reported to a rabies vaccination centre (duplicates excluded) [E]
Total under-reporting [1-E/B] (MF¼A/E)
166
153 92.17% (7.83%)
169
Rural health centre/ Health houses General practitioner in private sector Specialist practitioner in private sector Hospital
149 88.16% 0
130
12.75%
133
13.07% (1.25)
20 11.83%
e
0
e
16
20%
133
CDSS, communicable diseases surveillance system; MF, multiplication factor.
0
0
p u b l i c h e a l t h 1 5 4 ( 2 0 1 8 ) 1 3 0 e1 3 5
Rural health centre/ Health houses General practitioner in private sector Specialist practitioner in private sector Hospital
Number Reported to Under-reporting Reported cases Reported to Under Total of visits county's to the next to county's health the highest -reporting under-reporting [C] health level [1-D/C] centre (duplicates level [F] to the [1-F/B] (MF¼A/F) centre [D] excluded) [E] highest level [1-F/E]
134
p u b l i c h e a l t h 1 5 4 ( 2 0 1 8 ) 1 3 0 e1 3 5
Table 3 e Timeliness of Iranian CDSS for leishmaniasis and brucellosis. Number of participants
Number of detected cases
Cases reported to the highest level
Time between the onset of symptoms to the first presentation at a medical care centre (days) [Mean (med)]
Time between the first presentation at a medical care centre to being reported to the highest level of CDSS (days) [Mean (med)]
Total elapsing time (days) [Mean (med)]
Leishmaniasis
48,771
199
92
32.66 (29)
48.95 (30)
81.61 (63)
Brucellosis
48,771
55
32
34.5 (24)
22.0 (20.5)
56.5 (48)
Disease
Abbreviation: CDSS, communicable diseases surveillance system; med, median.
medical or healthcare provider. Only about 57% of the leishmaniasis cases who sought medical services were registered and reported to the highest level of CDSS. The average elapsing time for reporting leishmaniasis cases was about 80 days, which is too long to make CDSS able to implement preventive measures effectively. The results of the present study also suggested that all suspected patients of brucellosis sought diagnosis and treatment services and presented themselves to at least a medical centre. On the other hand, only 54.55% of the brucellosis cases who presented themselves to health or medical centres were reported to the higher levels. The elapsing time for brucellosis cases who sought health and medical services and were registered and reported to the highest level of surveillance system was too long (about 2 months) to make CDSS able to initiate control programs. Of all individuals who were bitten by an animal, 92.17% presented themselves to a medical centre. According to the results, 86.93% of the cases of animal bites were reported to the highest level of surveillance system and predominantly were vaccinated within less than 72 h since the animal bite. In spite of the relatively good sensitivity and timeliness of reporting animal bites, because of the fatal importance of ontime detection and vaccination of the cases, further improvement in quality of surveillance system is necessary. Regarding the high incidence rate of leishmaniasis, brucellosis and animal bites in Fars province, it seems that improvement in the timeliness and sensitivity of reporting cases is essential for effective controlling programs. It worth noticing that only a few studies have been conducted on the quality of the Iranian CDSS.5e7 The results of similar studies in the United States (2008, 2011) has shown that sensitivity of case reporting for brucellosis and rabies were 25% higher and 7% lower than the corresponding results of the present study respectively.8,10 Based on the results of another study in Greece (2010), sensitivity for
health houses. Of 153 cases who presented themselves to a health provider, 133 (86.93%) cases were registered and reported to the highest level of surveillance system. Accordingly, the rate of total under-reporting and MF of animal bites for CDSS was 13.07% and 1.25 respectively (Table 2).
Timeliness of surveillance system Leishmaniasis: The average elapsed time from the onset of symptoms to the first patient's presentation to a medical care centre was 32.66 days. The average elapsed time from patients' first presentation at the medical centres to the time of reporting cases to the highest level of surveillance system was 48.95 days. Accordingly, the total elapsing time between the onset of symptoms and reporting cases to the highest level of the health care system was on average 81.61 days (Table 3). Brucellosis: The average elapsed time between the onset of symptoms of brucellosis and the first patients' presentation to a medical care centre was 34.5 days. The average elapsed time between patients' first presentation to medical care centres and reporting cases to the highest level of surveillance system was 22 days. Accordingly, the average total elapsing time between the onset of symptoms and reporting cases to the highest level of the surveillance system was 56.5 days (Table 3). Animal bites: Only 84.96% of the detected cases were vaccinated within 24 h since animal bite. About 9.02% of the cases were vaccinated within 24e72 h and 6.02% received vaccination after 72 h from the date of animal bite (Table 4).
Discussion The results of this evaluation study suggested that more than 19% of symptomatic leishmaniasis cases did not visit a
Table 4 e Timeliness of Iranian CDSS for animal bites (rabies). Number of participants 48,771
Number of detected cases
Number referred to rabies vaccination centres
166
133
Abbreviation: CDSS, communicable diseases surveillance system.
The elapsing time between animal bite and vaccination [n (%)] <24 h 113 (84.96)%
24e72 h
>72 h
12 (9.02%)
8 (6.02%)
p u b l i c h e a l t h 1 5 4 ( 2 0 1 8 ) 1 3 0 e1 3 5
both brucellosis and leishmaniasis reported at about 74%, which is significantly higher than the corresponding rate obtained for Iranian CDSS.11 Strength and weakness: The present study was conducted in rural areas of a large province in the southern part of Iran (Fars province). To calculate the study indices, a populationbased door-to-door case finding was conducted to obtain required information independent to the CDSS database. No clinical information was obtained from medical and health centres that the patients visited during their course of disease. As the result, only symptomatic cases were detected and no attempt was made to confirm the status of diseases among the participants.
Conclusion Based on the results of the present study, the quality of case reporting for notifiable zoonotic diseases is relatively low. Disease surveillance indices must be routinely measured and improved at both governmental and private sectors. Training health personnel, especially physicians, about the objectives of the surveillance system and its importance on a periodic base is necessary. Routine evaluation of the performance of CDSS can effectively improve the validity and efficiency of the system.
Author statements Ethical approval This study was approved by the Shiraz University of Medical Sciences ethical review board.
Funding The present study was financially supported by Shiraz University of Medical Sciences, Shiraz, Iran.
135
Competing interests None declared.
references
1. German RR. Updated guidelines for evaluating public health surveillance systems. MMWR Recomm Rep 2001;50(1e35). 2. Organization, W.H. Communicable disease surveillance and response systems: guide to monitoring and evaluating. 2006. 3. Organization, W.H. International health regulations (2005). World Health Organization; 2008. 4. Heymann DL. Control of communicable diseases manual. American Public Health Association; 2008. 5. Fararouei M, et al. National guidelines for outbreak investigation: an evaluation study/Lignes directrices re d'investigation des flambe es: une e tude nationales en matie valuation. East Mediterr Health J 2013;19(9):816. d'e 6. Nejat M, et al. Surveillance of zoonotic diseases in Marvdasht city, Fars province. J Health Sci Surveill Syst 2016;4(2):95e102. 7. Fararouei M, Marzban M, Shahraki G. Completeness of cancer registry data in a small Iranian province a captureerecapture approach. Health Inf Manag J 2016. 1833358316668605. 8. Sickbert-Bennett EE. Completeness of communicable disease reporting, North Carolina, USA, 1995e1997 and 2000e2006Volume 17, number 1dJanuary 2011. Emerg Infect Dis J 2011. 9. Gibbons CL, Mangen M-JJ, Plass D, Havelaar AH, Brooke RJ, Kramarz P, et al. Measuring underreporting and underascertainment in infectious disease datasets: a comparison of methods. BMC Public Health 2014;14(1):147. 10. Thiede H, Close NS, Koepsell J, Baer A, Duchin JS. Completeness of reporting of rabies postexposure prophylaxis in King County, Washington. J Public Health Manag Pract 2008;14(5):448e53. 11. Jelastopulu E, Merekoulias G, Alexopoulos E. Underreporting of communicable diseases in the prefecture of Achaia, western Greece, 1999e2004-missed opportunities for early intervention. Euro Surveill 2010;15(21):19579.