Barefooted OR and Its Extension - Health System Planning in Rural Bangladesh Shams Rahman
Barefooted OR and Its Extension - Health System Planning in Rural
Bangladesh
Shams Rahman
Plan of the Presentation
• Background of the study
• Locational Analysis for Community Clinics
• Conclusion
• Q & A
4
1
2 3
44 4
1 1
22
3 3
Figure i: Successively-inclusive hierarchy Figure ii: Locally-inclusive hierarchy Figure iii: Successively-exclusive hierarchy
Symbol Facility Level Service Type
3
2
1
1, 2, 3
1, 2
1
3
2
3
2
Represent demand for service type 1
Represent demand for service type 2
Represent demand for service type 3
Health Systems
THC
HFWC
CC Mouza
Health Services Hierarchy at Thana level and below
Healthservices planning
Conceptualmaps
Locationalmodelling
Blending ‘Soft’ and ‘hard’ approaches
Education and knowledgeabout health and healthservices
Affordability(Cost/time)
Management of healthservices resources
Distribution/availabilityof static health facilities
Demand Supply
Conceptual map of health services planning problem
DEMAND SUPPLY
Education
BasicUnderstanding
of Health
Knowledge
Attitude
Religion/believes
TravelDistance(Time) Affordability
(Cost)
Availabilityof Doctors/Paramedics Availability
of Resources
Quality ofService
(Perceived)Quality of
Service(Potential)
Location
Electricity
Water Supply
PopulationSize
Growtharea/Hat
School/Post office
ExistingFacility
DABIR ALI (The Decision Maker)
To Take or Not to Take (Service)!
TransportSystem
Conceptual map - expanded
Developmentand Extension ofHealth Services
Government ofBangladesh
World Bank
European Commission (EC)
Policy
TFIPP (Pilot Project)
Implementation
MOHFW
Health policy development and implementation bodies
Policy Guidelines
► Government, in conjunction with the World Bank, developed the following policy guidelines:
Physical Accessibility:
Community Clinics (CC): Must be accessible within half an hour travel-time.
• Maximum travel distance (S) = 2 km
HFWC: One in each Union; S = 4 km (Total Union = 4500)
Policy Guidelines
►Population Coverage: One CC, Government suggested three service delivery options. These are:
One CC per 1500 population One CC per 6000 population, and One CC per 3000 population.
• One HFWC will serve a population of about 25000.
Policy Guidelines
►Potential sites:
Village size Uncapacited facility Electricity and acceptable water supplies, School
Locational efficiency: 7 existing plus 3 proposed and 7 existing plus 3 optimal
200
250
300
350
400
450
500
550
Number of facilities
Per
son
-kil
om
eter
s ('
000) Proposed
Optimal
7 8 9 10
0
20
40
60
80
100
120
1 2 3 4 5 6 7
Number of facilities (p)
% p
opul
atio
n co
vere
d
Optimal
Existing
0
20
40
60
80
100
120
7 8 9 10
Number of facilities (p)
% p
opul
atio
n co
vere
d
Optimal
Proposed
Findings and implications
► The Government's policy of locating one HFWC in each union and one CC per 6000 (considered the maximum level of population coverage) will make a health care delivery system consisting of 10 HFWCs and 40 CCs for the study area.
Findings and implications
► Geographically constrained
► Unconstrained problem
Findings and implications• Analysis was done locating 20 CCs (solved with the constraint of 2 in each
union) given the sites of the existing HFWCs. • The solution of this problem covers over 98% of the population, whereas the
solution of the unconstrained problem covers the entire population.
• The same quality of service (98%) can be obtained by optimally locating 13 CCs with the existing HFWCs.
• So, although the existing health delivery system (with 10 HFWCs) was found to be inefficient, an addition of 13 facilities optimally located would make the system comparable to the optimal system.
• Hence there are two solutions to the location problem which cover nearly all the population.
• One uses 22 facilities optimally located and the other has the 10 existing
facilities plus 13 more, optimally located. The difference in the population covered is only 0.6%.
Findings and implications (cont.)
• This solution is valuable for two reasons:
• First, implementation of such a solution would make a health delivery system 35% less costly (13 out of 20 CCs), while being very close to maintaining the objective of serving the entire population within a distance of 2 km.
• Secondly, the solution could automatically be used as a schedule for the physical opening of the facilities.
• Since it can be assumed that the Government would not have enough resources to finance the construction and equipping of 20 CCs at a time, the facilities will perhaps be opened over a period of time. The construction and equipping of 13 facilities would form the first phase. Interestingly, this would guarantee the optimality of the health deployment system even after the implementation of the second phase.
Findings and implications (cont.)
• In the second phase the health planner could open 6 or 7 further CCs if enough resources were available.
• To ask for resources for the construction and operation of 65% or 70% of the CCs (13 or 14 out of 20) might seem optimistic. However, it is only little over 32% of the government's allocation, of funds for 40 centres.
• The analysis has suggested several sets of solutions with respect to the location of community clinics and HFWCs for efficient delivery of health care in the region. The objective of the study was not to find a single optimal decision. Instead, we chose to develop and test feasible decision processes in the light of the Government's health policies. The results, therefore, should be regarded as an aid to the local health planner's intuition and not as a total replacement for it.
Concluding Remark
• What useful role can this study play?• By demonstrating that better options
(solutions) exist, the study provides community groups and bureaucrats/managers with a stronger case against inefficient and inequitable proposals.
Questions Please
Relationship between distance and diarrhoeal mortality
0
20
40
60
80
100
1 2 3 4 5 6
Distance (miles)
% patient's attendance
0
50
100
150
200
250
300
Mortality rateper 100,000
Attendance
Mortality
P-Median Problem
minimise a d xi ij ij
j
n
i
n
11
subject to: x iij
j
n
1
1,
x pij
j
n
1
xij xjj i j , , ; where i = j
xij 0 1,where
xij 1 if demand i is assigned to a facility j;
0 otherwise. n = number of demand points ai = population of demand i dij = shortest distance between i and j p = number of facilities to be located
Location Set Covering Problem (LSCP)
m i n i m i s e x j
j
n
1
s u b j e c t t o : x i nj
j N i
1 1 2, , , . . . . ,
x j nj 0 1 1 2, , , , . . . ,w h e r e
x j = 1 i f a f a c i l i t y i s l o c a t e d a t j ;
0 o t h e r w i s e .
N j d Si i j i / - s e t o f f a c i l i t y e l i g i b l e t o p r o v i d e s e r v i c e t o d e m a n d i
S = m a x i m a l s e r v i c e d i s t a n c e
Maximal Covering Location Problem (MCLP)
m a x i m i s e a yi i
i
n
1
s u b j e c t t o : x y i nj i
j N i
, , , . . . . ,1 2
x j pj
n
1
x j nj 0 1 1 2, , , , . . . ,
y j ni 0 1 1 2, , , , . . . ,w h e r e
x j = 1 i f a f a c i l i t y i s l o c a t e d a t j ;
0 o t h e r w i s e .
y i = 1 i f d e m a n d i s c o v e r e d b y a f a c i l i t y ;
0 o t h e r w i s e .
N j d Si i j i / - s e t o f f a c i l i t y e l i g i b l e t o p r o v i d e c o v e r t o d e m a n d
S = m a x i m a l s e r v i c e d i s t a n c e .
Trade-off Analysis
Population Coverage
0
20
40
60
80
100
120
1 2 3 4 5
No of facility
% C
over
ed
Trade-off Analysis
Population Coverage
0
20
40
60
80
100
120
1 2 3
No. of Facility
% c
over
ed
Trade-off Analysis
Population Coverage
0
20
40
60
80
100
120
1 2 3 4 5 6
No. of Facility
% C
over
ed