Page 91 - Contributed Paper Session (CPS) - Volume 5

CPS1108 Collins O. et al.
Table 1. General Descriptive Statistics of Maternal, Neonatal and Child Health
Data of Nairobi, Mombasa and Kisumu retrieved from DHIS2 for the period
Dec 2014 to Jan 2017
N Min Max Mean Std.
(monthly) (monthly) (monthly) Deviation
Live birth 60 49,810 87,329 72,711 9,263

Normal Deliveries 60 43,822 77,991 64,246 8,366
Neonatal deaths 60 406 1,038 804 169

Caesarean Sections 60 6,685 12,023 9,848 1,249

Assisted vaginal delivery 60 198 1,609 653 270
Babies discharge Alive 60 46,129 82,571 68,588 8,794
Maternal Deaths 20+ 60 45 337 84 36
years
Maternal Deaths 60 21 93 53 19
Audited
During the period December 2014 to January 2017, maternal complication
data from 896 facilities were retrieved with 48% of the facilities having either
basic emergency obstetric and new-born care (BEmONC) or comprehensive
emergency obstetric and new-born care (CEmONC) services.
Table 2. General Descriptive Statistics of Maternal Complications Data
retrieved from DHIS2 for the period Dec 2014 to Jan 2017
N Min Max Mean Std.
(monthly) (monthly) (monthly) Deviation
APH (Ante partum 60 231 773 485 100
Haemorrhage)
Eclampsia 60 204 533 393 79
Obstructed Labour 60 646 3001 973 290
PPH (Post-Partum 60 541 1260 902 185
Haemorrhage)
Ruptured Uterus 60 47 93 66 11
Sepsis 60 53 236 135 31

b. Time series Model
Let a count time series be Λt: t ϵ N and γt : t ϵ N be a time-varying r-dimensional
covariate vector, say γt = (γt,1, γt,2 , ···, γt,r). We can model the conditional

expectation E (Λt |ℐt-1) of the count time series by a process, λt : t ϵ N : E (Λt |ℐt-
1= λt). ℐt is the history of the joint process Λt, λt , γt : t ϵ N up to time t including
the covariate information at time t-1. The models' general form will be

80 | I S I W S C 2 0 1 9