Sunday, July 31, 2016

TIME SERIES ANALYSIS OF DENGUE CASES at BGHMC FROM YEAR 2011 to JULY 29, 2016

TIME SERIES ANALYSIS OF DENGUE CASES at BGHMC FROM YEAR 2011 to JULY 29, 2016

Purpose of the report: To describe the Time table and analyzed the series of dengue cases who were admitted and discharge at BGHMC
Scope of the cases: The data adopted in this study are patients who were diagnose to have contracted the dengue virus as primary diagnosis upon discharge from the year 2011 to July 29, 2016.

Table 1. Monthly discharges of dengue cases from Year 2011 to July 29, 2016
Year/Month
2011
2012
2013
2014
2015
2016
January
15
8
19
9
29
81
February
11
8
14
3
7
55
March
9
4
9
3
7
32
April
7
6
11
6
3
34
May
8
22
30
6
11
88
June
25
98
180
14
38
215
July
72
141
288
38
113
285
August
180
71
285
36
223

September
90
43
172
34
136

October
25
44
71
18
108

November
19
36
47
26
77

December
12
43
35
42
59

Total
473
524
1161
235
811
790

From the data obtained from the Mortality and Morbidity Spreadsheet shows that the peak cases were during the year 2013 from with 1161 cases, while the year 2016 have a total of 790 cases as of July 29, 2016 which have a high probability to increase on the succeeding months since it was observed in the table that June to November have high number of dengue cases usually during the wet season. The data might suggest that there is a seasonality of dengue cases every 3 years.  
However to describe the series of this dengue case time series analysis should be done to have an Idea about the trend, and the future increase of this dreaded disease as illustrated below.
Figure 1. Time series plot of Dengue Cases From year 2011 to July 29, 2016
Figure 1 shows the time series plot of dengue cases from 2011  to July 29, 2016 wherein we can see a lot of spikes which would suggest a seasonality per month, however some of the spikes or its peak are quite higher than the other months suggesting that unusual increase of dengue discharges. Those unusual graph where seen during the months of 31-34 and 65-67 which are on the months of July to October 2013 and May to July 2016. 

Due to irregular increase and decrease of dengue cases monthly it’s hard to forecast for the next few months. However smoothening and other techniques the Figure below give a rough forecast for the next 6 months.
From table 2 and figure 2 denotes an increase for the next 5 months if the data would have same flow or in normal circumstance. The peak may be seen during the month of October and November and hopefully would decline during the end of the month till the dry season.
Figure 2.Time Series Forecast for the Next Month


Table 2 Rough forecast of dengue cases
Forecast Value
From
To
Month
Year
129
265
August
2016
100
237
September
2016
132
270
October
2016
169
308
November
2016
148
288
December
2016

Conclusion:
Dengue fevercases have a significant increase this year comparing last year.
The viral disease might have yearly seasonal characteristics every three years aside from the monthly seasonality which often increases during the wet season.
The viral disease would continually increase until the end of the month unless there would be an intervention.

Darius T. Muñoz
HIMO (July 29, 2016)


All data needs consent to the author 

Tuesday, May 14, 2013

Comparison of Amniotic Fluid Gram Stain and Leukocyte Esterase Activity in the Prediction of Subclinical Intraamniotic Infection



Chorioamnionitis is frequently associated with prolonged rupture of membranes, prelabor rupture of membranes (PROM), preterm prelabor rupture of memebranes (PPROM), maternal urinary tract infection, prematurity and maternal fever.  Though may be absent, signs and symptoms of chorioamnionitis include fever >37.8°C, significant maternal tachycardia (>120 beats per minute), fetal tachycardia, purulent or foul-smelling amniotic fluid or vaginal discharge, maternal leukocytosis (15,000-18,000 cells/mm3) or uterine tenderness (Sherman and Otsuki 2003).
Several serologic tests can be done on the mother or tests on the amniotic fluid can confirm clinical suspicion of chorioamnionitis.  Amniotic fluid gram stain and culture is reliable in the setting of chorioamnionitis.  Detecting glucose levels in the amniotic fluid is also an indirect indicator of chorioamnionitis as glucose is consumed by the rapidly dividing bacteria.  Another indicator for infection is an elevation of cytokines.  C-reactive protein determination was found to have a sensitivity and specificity of 80% as an early predictor of subclinical chorioamnionitis (Saini et. al. 2003).  In a study done by Grable and Heine in 2003, maternal plasma levels of defensin, a neutrophil granule product, is 76% sensitive and 94% specific in predicting histologic chorioamnionitis. Other studies have related amniotic fluid glucose concentration below 5 mg/dl (Kiltz et. al. 1991), elevated matrix metalloproteinase (MMP)-9, interleukin (IL)-6, and IL-12 (Harirah et. al. 2002, McNamara 2003) as markers for subclinical chorioamnionitis.  Unfortunately, many of these markers are routinely available especially in a developing country like ours, thus the need for cheaper alternatives for early diagnosis of chorioamnionitis. http://onlinelibrary.wiley.com/doi/10.1111/j.1742-4658.2010.07919.x/abstract
            Berhman et. al. describes neonatal infections that manifested in the first week of life is usually attributable to microorganisms transmitted from mother to infant.  Epidemiologic studies have subdivided neonatal infections to two: 1) early-onset, occurring within the first 72hrs of life and is assumed to have been transmitted perinatally from the mother, and 2) late-onset, occurring between 3-7 days.  It is assumed that late-onset infections are acquired post-natally from an environmental source.  Microorganisms most commonly associated with early-onset infection include group B Streptococcus (GBS), Escherichia coli, Haemophilus influenzae, and Listeria monocytogenes.
Signs and symptoms of neonatal sepsis are often nonspecific and subtle. Behavioral abnormalities manifested by the neonate may include lethargy, hypotonia, weak cry, and poor suck. Specific organ involvement may be manifest by 1) Pulmonary - tachypnea, respiratory distress, cyanosis, hemorrhage, and/or apnea; 2) Cardiovascular - tachycardia, hypotension, cool and clammy skin, pale or mottled appearance, and/or oliguria; 3) Gastroinstestinal symptoms - abdominal distension, vomiting, diarrhea, and/or bloody stools; 4) CNS - thermal regulatory abnormalities (hypothermia or hyperthermia), behavioral abnormalities, apnea, and/or seizures; 5) Hematopoietic system - pallor, petechiae or purpura (Anderson-Berry et. al. 2006)


Between the two simple and rapid tests, gram stain of amniotic fluid was superior to leukocyte esterase activity as predictors of subclinical intraamniotic infection in terms of higher sensitivity, positive and negative predictive values.
Although leukocyte esterase activity and gram stain had positive correlation with the detection of subclinical intraamniotic infection (p value 0.012 and 0.000 respectively), results are insufficient to be a basis for decision making.  Clinical signs and symptoms are still needed for proper evaluation and management of patients.
    Among patients positive for leukocyte esterase activity and microorganisms on gram stain, it would be prudent to provide mothers with extra education to be more observant with their babies on the development of subtle signs of infection. Likewise both pediatricians and obstetricians are advised to be more vigilant in monitoring these patients to prevent morbidities