SEARCH

SEARCH BY CITATION

Keywords:

  • schistosomiasis;
  • integrated control;
  • passive case finding;
  • health system;
  • Ghana

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Passive case finding based on adequate diagnosis and treatment of symptomatic individuals with praziquantel by the health care facilities is a minimum requirement for integrated schistosomiasis control. Two field studies were conducted in Ghana to obtain quantifications about the steps in this process: (1) a study of health-seeking behaviour through interview of individuals with reported schistosomiasis-related symptoms; (2) a study of the performance of the Ghanaian health system with regard to schistosomiasis case management by presenting clinical scenarios to health workers and collecting information about availability of praziquantel. It appeared that cases of blood in urine (the most typical symptom of Schistosoma haematobium) and blood in stool (the most typical symptom of S. mansoni) have a very small probability of receiving praziquantel (4.4% and 1.4%, respectively) from health facilities. Programmes aimed at making the drug available at all levels of the health care delivery system and encouraging health-seeking behaviour through health education are not likely to increase these probabilities beyond 30%. This is because many cases with blood in urine do not consider it serious enough to seek health care, and blood in stool usually requires (imperfect) diagnostic testing and referral. We therefore conclude that additional control activities, especially for high-risk groups, will remain necessary.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Over the past decades, control of urinary and intestinal schistosomiasis was mainly based on vertical programmes, co-ordinated and organized at the national level and usually supported by donor organizations (e.g. El Malatawy et al. 1992; Barakat et al. 1995). The aim was to reduce transmission and infection in populations to levels at least low enough to minimize the risk of serious morbidity. As a result of the relatively high cost of praziquantel, screen-and-treat campaigns (selective population chemotherapy) were considered the most cost-effective strategy in most endemic countries (World Health Organization 1985). However, the long-term results were often disappointing because of rapid re-infection and the expensive nature of these programmes that make them unsustainable after withdrawal of external funding (Gryseels 1989; Kumar & Gryseels 1994).

In 1993, the WHO Expert Committee called for a more prominent role of the regular Primary Health Care System (WHO 1993). Building blocks for this integrated control were defined as health education, diagnosis and treatment, promotion of safe water supply, sanitation and snail control. It was, however, emphasized that the first essential component should be adequate clinical care for patients presenting at a health post or clinic with early signs and symptoms (passive case finding). In any case, this is the only option in countries without any form of organized schistosomiasis control (Engels et al. 2002). As pathology is strongly related to intensity and duration of infection, treating early cases may also prevent most of the severe morbidity later (WHO 2002). Moreover, such treatment meets the demands of the populations and strengthens the health system as a whole. The recent resolution of the 54th World Health Assembly (WHA 54.19, 22 May 2001) and the WHO Expert Committee on schistosomiasis and soil-transmitted helminthiasis (WHO 2002) call for a comprehensive approach and recommend ensured access to treatment in primary health care services, associated with regular delivery of treatment to high-risk groups, particularly school-age children, and implementation of plans for basic sanitation and safe water supplies.

The prerequisites for the success of integrated schistosomiasis control through passive case finding are adequate health-seeking behaviour, good access to functional health care facilities, proper treatment and referral strategies and availability of praziquantel. In a multidisciplinary research programme, we have tried to obtain quantifications for each of these steps (Figure 1) by conducting field studies in West Africa. This paper aggregates our major findings for Ghana, a country currently restructuring its health care delivery system by shifting responsibilities to peripheral health facilities, with integration of control of parasitic diseases – including schistosomiasis – in the regular health services and active community participation as central components. We apply the quantifications from our study to estimate the probability that individuals with schistosomiasis-related symptoms receive praziquantel from the Ghanaian health system, and we explore the impact of improvements on this process.

image

Figure 1. Schematic representation of prevention of schistosomiasis morbidity by treatment of cases reporting with mild symptoms.

Download figure to PowerPoint

Methods

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

The data for this study were based on two field studies reported elsewhere: one about health-seeking behaviour of individuals with schistosomiasis-related symptoms and the other about functioning of the health system with regard to schistosomiasis control. A brief description of each study is given below.

A questionnaire-based study was conducted in Kokoetsekope (Greater Accra Region, Ghana) from July to September 2000. This rural village is located along the Densu Lake and has about 380 inhabitants, most of them with no or only basic, elementary education. A preliminary survey by the Noguchi Memorial Institute for Medical Research involving a random sample of a small portion of the population resulted in prevalences of Schistosoma haematobium (by urine filtration) and S. mansoni (by Kato-Katz faecal smear) of 70% and 78%, respectively. In Kasoa, a nearby commercial town about 2 km away, two health centres and 11 private clinics were available for conventional medical treatment. The health centres are government owned and manned by medical assistants. Patients were required to make full payment for consultation and laboratory investigation (according to the cash and carry system) before treatment was provided. The private clinics were headed by either medical doctors or nurses. All inhabitants were interviewed, and for those younger than 6 years or who could not answer the questions themselves, a parent or guardian responded for them. They were asked if they had experienced any of seven schistosomiasis-related signs and symptoms (blood in urine, painful urination, diarrhoea, blood in stool, swollen abdomen, abdominal pain, fatigue) and fever within 1 month before the day of interview. Individuals who reported a sign/symptom were asked what action was taken and whether this included a visit to a health care facility (health centre or clinic). In case they had not visited a health care facility for the reported sign/symptom, they were asked why not. In the present study, we mainly focussed on the results reported for the symptoms blood in urine and blood in stool (including bloody diarrhoea). More details of this study and results for other signs/symptoms can be found in Danso-Appiah et al. (2004).

In April and May 2000, a random sample of 70 health care facilities (12 hospitals, 53 health centres and mission clinics, and five private clinics) in four geographically different areas of Ghana were visited. The person in charge of the health facility was interviewed without revealing our specific aim of evaluating schistosomiasis control. We started by presenting four clinical scenarios with symptoms related to schistosomiasis and asked the respondents to explain their usual case management policy for such patients. We wanted to find out if symptom-based treatment was performed, if diagnostic testing was requested and if there was referral for a diagnostic test or treatment. We further asked what the action would be after a positive (i.e. identification of Schistosoma eggs) and a negative test, and which treatment would be prescribed. Thereafter, we revealed the focus of our study and asked whether S. haematobium or S. mansoni infection were present in the coverage area. Respondents from areas with reported schistosomiasis were interviewed using a structured questionnaire about their knowledge of schistosomiasis symptoms and availability and costs of diagnostic tests and praziquantel. In the present study, we used the results for both clinical scenarios most distinctive for urinary and intestinal schistosomiasis, respectively: a 10-year-old girl with blood in urine, without any other signs and symptoms; and a 10-year-old boy with abdominal discomfort and bloody diarrhoea without any other signs and symptoms.

Probabilities of receiving a prescription of praziquantel were calculated from all situations where praziquantel was prescribed directly, after positive diagnostic testing (assuming a 50% chance of a positive result), or after referral of patients to hospitals (using the average management policy of all hospitals in the study, and assuming that all patients will comply with referral and buy praziquantel). The 50% chance of a positive result was based on literature review and appeared a reasonable estimate for both the proportion of patients with blood in urine showing S. haematobium eggs in a standard urine centrifugation test, and the proportion of patients with blood in stool showing S. mansoni eggs in a standard direct smear test (other, more sensitive tests were not reported; see Van der Werf et al. 2003, 2004, for more details and results of this field study).

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

There were 318 inhabitants of Kokoetsekope present at the time of the study and only one refused to participate. Using a 1-month recall period, 64 individuals reported blood in urine and 19 (30%) of them reported to have taken action. Nine (47%) of the 19 cases visited a health care facility as first action. Blood in stool was reported by 82 individuals, of whom 48 (59%) took action, among which 12 (25%) visited a health centre or clinic as first option. Another action mostly taken was self-medication, with allopathic drugs four to five times more often used than herbal treatment. The overall proportion of cases visiting a health care facility (about 14% for both symptoms) was lower than for fever (30%) and slightly lower than the average for the other schistosomiasis-related signs/symptoms (about 20%). Overall, teenagers (10–19 years) with schistosomiasis-related symptoms showed a significantly lower tendency to visit a clinic or health centre than children and adults (9%vs. 21%).

All 70 selected health care facilities agreed to participate in our study. Interviews of the persons in charge revealed that clinical scenarios comparable to the one about blood in urine have a chance of 66% to receive a prescription of praziquantel, whereas this was 13% for the scenario about blood in stool. For both scenarios, prescription was usually made after diagnostic test (assuming 50% chance of positive result) or it was made in a hospital in case of a referral. The low probability of receiving praziquantel for the second scenario was largely because schistosomiasis was rarely considered as the most likely initial diagnosis (3% when compared with 91% for the first scenario). The proportions of health facilities among those prescribing praziquantel that had it in stock were 47% and 74% for the first and second scenario, respectively. The difference is mainly due to the fact that hospitals (which usually have praziquantel in stock) constitute the largest part of the few health facilities that prescribe praziquantel in case of blood in stool.

Table 1 summarizes the successive steps between having blood in urine or blood in stool and receiving praziquantel, using above quantifications. The overall probability to receive praziquantel from the health system is 4.4% for cases with blood in urine and 1.4% for cases with blood in stool.

Table 1.  Successive steps in the probability to receive praziquantel for cases with blood in urine and blood in stool
StepBlood in urine (%)Blood in stool (%)
  1. Proportion seeking health care and visiting a clinic or health centre were based on 64 (of 317 interviewed) individuals in the village of Kokoetsekope (Greater Accra Region, Ghana) who reported blood in urine and 82 individuals who reported blood in stool (including bloody diarrhoea) within one month before the day of interview. Probability of health care facilities prescribing praziquantel and proportion having it in stock were based on interview of health workers in a random sample of 70 health care facilities in different geographical regions of Ghana.

1. Proportion seeking health care3059
2. Proportion visiting a clinic or health centre out of those seeking health care4725
3. Probability of receiving a praziquantel prescription6613
4. Proportion of health care facilities having praziquantel in stock of those prescribing it4774
Overall probability 4.4 1.4

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

The process of passive case finding as an essential part of integrated disease control has been investigated before, but mainly in a qualitative way. The TB model by Waaler and Piot (1969) is the best known example. To our knowledge, we are the first to have conducted field studies within one country to obtain quantitative information about the steps from perceived symptoms to receiving proper treatment. Although schistosomiasis seems a classical example for integrated disease control (WHO 1993), our study revealed a rather disappointing result for passive case finding in Ghana, with an overall probability of <5% cases with blood in urine or blood in stool receiving praziquantel from the health system. The main bottleneck for blood in urine is the low tendency of symptomatic cases to seek health care. For blood in stool, and probably other symptoms caused by S. mansoni, the bottleneck is the low alertness of health workers about intestinal schistosomiasis, leading to a limited chance of prescribing praziquantel. The overall probability of receiving praziquantel (within or outside the health system) is slightly higher as the drug can sometimes be obtained from privately owned chemical shops (Danso-Appiah et al. 2004).

We believe that our measurements and calculations give a realistic picture of the situation in Ghana. Our study on health-seeking behaviour of subjects with schistosomiasis-related signs and symptoms was only a pilot study and constituted a relatively small community. However, preliminary analysis of data from a similar study in Ghana involving big samples of individuals from three locations of varying cultures and ethnicities did not show results markedly different from those presented here. Our study of the performance of health care facilities concerned a representative sample of endemic areas in Ghana, and the main results did not differ between the four regions visited. There was no particular reason for the interviewed health workers to give socially desirable answers, as our interest in schistosomiasis case management was revealed only after the clinical scenarios were presented. Also, the reported presence of praziquantel was cross-checked by the interviewer. Finally, both studies were characterized by a participation rate of nearly 100%.

It is striking that all four steps in Table 1 contribute to the low overall probability of receiving adequate treatment. This means that interventions aimed at improving only one step will not have a substantial effect on the overall outcome. Thus, interventions that have effect on multiple steps are necessary.

In this respect, making praziquantel available at all levels of the health system is the most logical option for improvement. Apart from a maximum availability of the drug (step 4), it may increase the probability of prescription (step 3). First of all, there seems more reason to prescribe the drug if it is directly available from the health care facility visited. In Mali, another country where the performance of the health care facilities with respect to schistosomiasis was investigated by our research team (Landouréet al. 2003), 85% of the respondents reported to prescribe praziquantel in case of blood in urine, mostly without diagnostic testing or referral (Van der Werf et al. 2004). The same study showed that a case with blood in stool in Mali still has a small chance to receive a prescription of praziquantel (19%), although 81% of all health centres had the drug in stock and most health workers knew this symptom of S. mansoni. This large difference is mainly due to difficulties in diagnosis of S. mansoni infection. Most health workers mentioned other diseases such as bacterial infection and amoebiasis when asked for the first diagnosis of a patient reporting with blood in stool, which may – from a public health point of view – very well be the best strategy based on the presence of other infections relative to S. mansoni. Only in the recent focus of Northern Senegal, with extreme levels of S. mansoni infection, diagnostic algorithms were developed and applied to prescribe praziquantel based on symptoms only (Van der Werf et al. 2002). In moderately endemic situations, some diagnostic testing will always be necessary. As most Ghanaian health centres and clinics at the peripheral level have no laboratory facilities, this is only possible at the hospital level. As a consequence, this entails the risk of loss of patients due to extra cost (diagnosis, travel, etc.) and time lost.

Availability of praziquantel may also increase the willingness of symptomatic cases to visit a health care facility (step 2). However, it is interesting to note that very few symptomatic cases expressed a negative attitude towards health care (4% reported ‘drugs do not help’) when asked about the reason for not going to a health centre or clinic (Danso-Appiah et al. 2004). Financial reasons (‘do not have the money’) and perceived severity of the disease (‘not serious enough’) appeared to be much more important (47% and 44%, respectively). A reduction of the cost of treatment may further increase the tendency to visit a health care facility, but it is not likely to attract the entire 47% as the cost of travel and the loss of income due to time lost (half a day to one day) in seeking hospital care will remain unchanged. Finally, socioeconomic status did not have any effect on the tendency to seek health care, except for the reported reason for not going to a health care facility (Danso-Appiah et al. 2004).

The steps in Table 1 can now be used to explore what would happen in an ideal situation of wide availability of cheap praziquantel (say 90% of health centres/clinics and 100% of hospitals) and training of health personnel about when to prescribe it. A policy of direct treatment for all cases reporting with blood in urine, assuming a small increase in visiting of health facilities, would lead to an overall probability of receiving praziquantel equal to 16% (step 1 × step 2 × step 3 × step 4 = 30%× 60% × 100%× 90%). A policy to first refer all reporting cases with blood in stool for diagnosis with Kato-Katz (about 70% sensitivity), assuming a small increase in visiting of health facilities but a 30% loss of patients due to referral, will lead to an overall probability of about 9% (59% × 30% × 50% × 100%). These probabilities are four to six times the original values, but still rather low.

Health education can further increase the probability of receiving praziquantel. It may raise awareness in the population about schistosomiasis-related symptoms (step 1), and also encourage symptomatic cases to go to their local health centre or clinic for proper treatment (step 2). In practice, however, it is difficult to put such messages through as long as populations at risk of schistosomiasis remain poor and have to deal with so many other infections. A study in Northern Senegal showed that even 7 years of health education as part of intense control and research activities were not enough to make more than half of the population accurately quoting symptoms associated with intestinal schistosomiasis (Sow et al. 2003). Also, it will not be easy to change local attitudes and perceptions that have evolved over centuries of living with schistosomes: blood in urine is considered a fact of life in many areas endemic for urinary schistosomiasis and sometimes even seen as a sign of maturity (Asenso-Okyere et al. 1998). It is not likely that the number of cases seeking health care will change much, but health education may substantially increase the proportion of those that will go to the regular health care (say to 80%). The overall probability to receive praziquantel for cases with blood in urine will then be 29% (40% × 80% × 100% × 90%). For blood in stool, this is 26% (65% × 80% × 50% × 100%).

Thus programmes aimed at making praziquantel available at all levels of health care and encouraging health care seeking by health education are not likely to increase the probability of symptomatic cases receiving adequate treatment beyond 30%. This is because the most common mild symptoms (in particular blood in urine) are often not considered serious enough to seek health care, and blood in stool usually requires (imperfect) diagnostic testing and referral to specialized centres with the risk of loss of patients. A substantial number of cases with mild symptoms will therefore remain untreated with the risk of developing severe morbidity.

With respect to the recent WHO recommendations for schistosomiasis control, we can safely conclude that the current poor availability of praziquantel in health facilities (our study), together with the limited presence of other control measures (mainly confined to Volta Region), poses considerable challenges to the Ghanaian health system. To meet the specific WHO target for all Member States in endemic areas ‘of attaining a minimum target of regular administration of chemotherapy to at least 75% and up to 100% of all school-age children at risk of morbidity by 2010’ (World Health Organization 2002) will require huge efforts. Our conclusion that even with maximum access to praziquantel about 70% of episodes with blood in urine or blood in stool will be left untreated, demonstrates that integrated control by passive case finding cannot be the only component of successful morbidity control in endemic communities. This confirms current WHO policy that ensured access to treatment in primary health care services should always be complemented by other control activities, notably regular treatment of high-risk groups.

The current study presents a simple but unique approach to provide real-life quantifications to the process of health care integrated schistosomiasis control. Within the framework of the multidisciplinary research programme (illustrated in Figure 1) of which this study is part, we aim at conducting further research to elucidate the overall impact of treatment on reduction of schistosomiasis morbidity. For example, it is interesting to find out to which extent even occasional periods of reduced intensity of infection for those who received treatment after self-reporting with mild systems to the health system will regress, or at least halt, the development of severe pathology and morbidity. It can be imagined that passive case finding selects those with the highest risk. Similarly, more subtle consequences such as growth retardation and impaired cognitive development may be prevented by such occasional treatment of episodes with mild symptoms (Nokes & Bundy 1994). A final judgement of the consequences of integrated control by passive case finding or other treatment approaches can only be obtained after lengthy and costly cohort studies (e.g. Hatz et al. 1998). Simulation models aggregating our quantifications on health care seeking behaviour and functioning of health systems with respect to schistosomiasis control, together with an appropriate representation of the dynamics of progression and regression of pathology and morbidity, may be a good alternative.

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

The study was financially supported by the Netherlands Foundation for the Advancement of Tropical Research (WOTRO). The field work on the performance of health systems received additional support from the Foundation Vereniging Trustfonds Erasmus Universiteit Rotterdam. The Noguchi Memorial Institute for Medical Research (NMIMR) has offered the necessary human resources throughout the field studies. We thank all NMIMR field assistants who helped in the fieldwork and the inhabitants of Kokoetsekope for their kind co-operation.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  • Asenso-Okyere WK, Anum A, Osei-Akoto I & Adukonu A (1998) Cost recovery in Ghana: are there any changes in health care seeking behaviour? Health Policy and Planning 13, 181188.
  • Barakat R, El Masry AG, Farghaly A et al. (1995) Impact of population-based selective chemotherapy on prevalence and intensity of Schistosoma mansoni infections in the Nile Delta: Kafr El Sheikh. Tropical and Geographical Medicine 47, 266270.
  • Danso-Appiah A, De Vlas SJ, Bosompem KM & Habbema JDF (2004) Determinants of health seeking behaviour for schistosomiasis-related symptoms in the context of integrating schistosomiasis control within the regular health services in Ghana. Tropical Medicine and International Health 9 (in press ).
  • El Malatawy A, El Habashy A, Lechine N, Dixon H, Davis A & Mott KE (1992) Selective population chemotherapy among schoolchildren in Beheira governorate: the UNICEF/Arab Republic of Egypt/WHO Schistosomiasis Control Project. Bulletin of the World Health Organization 70, 4756.
  • Engels D, Chitsulo L, Montresor A & Savioli L (2002) The global epidemiological situation of schistosomiasis and new approaches to control and research. Acta Tropica 82, 139146.
  • Gryseels B (1989) The relevance of schistosomiasis for public health. Tropical Medicine and Parasitology 40, 134142.
  • Hatz CF, Vennervald BJ, Nkulila T et al. (1998) Evolution of Schistosoma haematobium-related pathology over 24 months after treatment with praziquantel among school children in Southeastern Tanzania. American Journal of Tropical Medicine and Hygiene 59, 775781.
  • Kumar V & Gryseels B (1994) Use of praziquantel against schistosomiasis: a review of current status. International Journal of Antimicrobial Agents 4, 313320.
  • Landouré A, Van der Werf MJ, Traoré M & De Vlas SJ (2003) Evaluation of case management in the integrated schistosomiasis-control programme in Mali. Annals of Tropical Medicine and Parasitology 97, 723736.
  • Nokes C & Bundy DAP (1994) Does helminth infection affect mental processing and educational achievement? Parasitology Today 10, 1418.
  • Sow S, De Vlas SJ, Mbaye A, Polman K & Gryseels B (2003) Low awareness of intestinal schistosomiasis in northern Senegal after 7 years of health education as part of intense control and research activies. Tropical Medicine and International Health 8, 744749.
  • Van der Werf MJ, Mbaye A, Sow S, Gryseels B & De Vlas SJ (2002) Evaluation of staff performance and material resources for integrated schistosomiasis control in Northern Senegal. Tropical Medicine and International Health 7, 7079.
  • Van der Werf MJ, Bosompem KM & De Vlas SJ (2003) Schistosomiasis control in Ghana: case management and means for diagnosis and treatment within the health system. Transactions of the Royal Society of Tropical Medicine and Hygiene 97, 146152.
  • Van der Werf MJ, De Vlas SJ, Landouré A, Bosompem KM & Habbema JDF (2004). Measuring schistosomiasis case management of the health services in Ghana and Mali. Tropical Medicine and International Health 9, 149157.
  • Waaler HT & Piot MA (1969) The use of an epidemiological model for estimating the effectiveness of tuberculosis control measures. Sensitivity of the effectiveness of tuberculosis control measures to the coverage of the population. Bulletin of the World Health Organization 41, 7593.
  • World Health Organization (1985) The Control of Schistosomiasis: Report of WHO Expert Committee. Technical Report Series No. 728, Geneva.
  • World Health Organization (1993) The Control of Schistosomiasis. Second Report of the WHO Expert Committee. Technical Report Series No. 830, Geneva.
  • World Health Organization (2002) First Report of the Joint WHO Expert Committees on the Prevention and Control of Schistosomiasis and Soil-transmitted Helminthiasis. Technical Report Series No. 912, Geneva.