OpenIDEO is an open innovation platform. Join our global community to solve big challenges for social good. Sign Up / Login or Learn more

diagnosis app

"EBMApp" is a non-profit internet-app for the diagnosis of Ebola. EBMApp would have diagnosed Ebola on patient zero in USA (Mr. Duncan) even before his 1st hospital visit. Check out the demo-video at www.ansay.de/demo

Photo of Can Ansay
0 0

Written by

EBMApp diagnoses Ebola
 
  • immediately, reliably, worldwide, lab-, risk- and cost-free

through
 
  • internet, hotlines or healthcare-/airport-staff.

EBMApp could be the game changer, Wendy Taylor (director of the Center for Accelerating Innovation and Impact at USAID) desired:
 
  • “Right now we have diagnostics, but they are lab-based diagnostics, and they take time. Having rapid point-of-care diagnostics could really be a game changer.”

Test EBMApp at www.ebmapp.org
 
  • Username: usaid@ebmapp.org  
  • Password: usaid

EBMApp is most useful when a possible Ebola patient calls a call-center in or outside Africa for diagnosis and the call-center agent then uses EBMApp on a PC. Toll-free hotlines are already in service, cellphones are widespread and the low literacy rate doesn´t matter on the phone. I´m already in contact with Tina J. Kpan, supervisor of the call-center in Liberia, the hardest hit country in West-Africa.
 
EBMApp could accomplish all 3 missions of the challenge:
 
Mission 1:
EBMApp is an enhancing tool in the field to empower and protect healthcare workers. With EBMApp healthcare workers can easily diagnose Ebola and narrow down the few uncertain cases which need Lab-tests. Infection risk will be reduced since number of dangerous Lab-test is reduced and diagnoses are mainly made through hotlines. Insolation wards will not be overcrowded anymore.
 
Mission 2:
EBMApp ensures that communities receive and respond to timely, accurate and actionable information about Ebola prevention and treatment. An expert-team will enter and update all data and information in EBMApp´s database. Therefore an Ebola-patient using EBMApp will get the best recommendation on treatment and isolation measures.
 
Mission 3:
EBMApp is a “mobile solutions to facilitate data collection and analysis”. EBMApp works browser-based on Laptop PCs, tablets and smartphones (Samsung and LG recently donated 5.000 smarthones). Each diagnosis can be saved on the server and be used for further research. Anonymous live surveillance of users´ diagnoses keeps track of Ebola.
 
I. How it works
1. Generally

"EBMApp" is the abbreviation of „ Evidence- Based Medicine Application”. EBMApp is based on a sophisticated algorithm (patent pending) developed in cooperation with top-level experts and therefore deeply "grounded in science", as president Obama desired.

Check out the 2nd additional demo-video at www.ansay.de/demo-long
or the video about how to use EBMApp at www.ansay.de/howto.

EBM App is fully functional only for the 2 main symptoms
 
  1. fever“ and
  2. pain > muscles > on whole body“. 
     
Since these 2 symptoms statistically occur at 96% resp. 79% of all Ebola patients, the combination covers more than 99% of all Ebola patients.
 
When you enter a main symptom on the upper-left side of the start-page, a list of all possible diseases appears in the middle column incl. death-risk, sorted esp. by incidence.
 
At the same time in the left column a list (sorted by relevance) appears with ca. 300 queries to all symptoms resp. other indicators, which statistically might occur in these diseases. The mouse-over-effect on each symptom-query highlights the corresponding disease(s) in the middle column.
 
According to which answer you give ([+] or [-]) the possibilities change until the most probable disease is determined with sufficient certainty. The probability of a disease increases the more
 
  • the more frequent the disease generally occurs (“incidence”),
  • the more frequent a present symptom generally occurs in this disease (“sensitivity”) and
  • the less frequent a present symptom occurs in other diseases (“specificity”). 
     
(By the way, you can view these evidence-data for each disease via a link on the right column.)
 
2. Example: 1st case of Ebola in USA
For example, if you enter “fever” as main symptom, Ebola appears on rank 42 on the list with a probability of 0,00006% (=6E-5), which becomes visible when you move the mouse-pointer over the displayed “<1%”.
 
For testing purposes click “[+]” on the queries regarding all the following symptoms, which occurred in the 1st Ebola patient in USA (Thomas Eric Duncan) when he first visited the hospital in Dallas:
 
  • fever, abdominal pain, dizziness, headache, decreased urination[1].
     
EBMApp shows the correct diagnosis: 100% Ebola.
 
EBMApp doesn´t even have to know that Mr. Duncan was in West-Africa and had contact to an Ebola patient.
 
The doctors however diagnosed sinusitis as well as abdominal pain and sent him home with a prescription for useless antibiotics.
 
Prior to this CT-scans and blood-tests were conducted. One of the results was “low counts of white blood cells & platelets”, an obvious clue for Ebola[2].
 
This result of the blood-test itself would have raised Ebola to 1st rank on EBMApp. Because the probability always immediately increases to 100% when extremely relative-specific symptoms or signs like lab-tests are present. But even symptoms with low specificity differentiate well, in case these symptoms statistically never occur in Ebola or competing diseases. Because if a symptom is present which statistically never occurs in Ebola, Ebola is definitely excluded with 0%.
 
3. Restrictions
All symptoms marked with an asterisk [*] can not yet be answered with [+] since the necessary data is not yet entered into the database. That does not matter much, because usually there are only a few symptoms present anyway. Besides the power of EBMApp to differentiate is mainly based on the info, which symptoms are absent although statistically they are usually present in certain diseases. Due to excluding these diseases the probability of the remaining diseases passively increases. EBMApp assumes no liability.
 
II. Best use
To use EBMApp in the best way answer the first 10-35 symptom-queries in the left column. Then you answer all queries regarding Ebola (marked with “<E”) and then regarding every disease, which ranks higher than Ebola or up to 3 ranks lower than Ebola. Regarding Ebola there are 3 simple results possible with corresponding diagnoses and advices:
 
  1. Probability of Ebola: by far rank 1
    > Diagnosis: Ebola [+]
    > Advice: Isolation and treatment according to recommendations in right column.
     
  2. Probability of Ebola: > 1%
    > Diagnosis: Ebola [?]
    > Advice: Isolation and Lab-test.
     
  3. Probability of Ebola: < 1%
    > Diagnosis: Ebola [-]
    > Advice: use EBMApp again if symptoms chance. 
     
The probability of Ebola is usually essentially more likely, if the patient
 
  • visited a region with a current outbreak or
  • had contact to Ebola-patients. 

III. Diagnosis of Ebola
EBMApp thus can at least reliably diagnose, if Ebola is present or not or if a Lab-test is necessary.
 
IV. Diagnosis of other diseases
In case Ebola is not present, EBMApp furthermore can diagnose, which disease is present instead, but with a lesser reliability. Because admittedly the list of possible diagnoses of the main symptom “fever” covers all possible diseases (ca. 90) and the other main symptom “pain > muscles > on whole body” covers the most common and deadly diseases (ca. 80 of 200).
However, rare or harmless diseases in this list are not yet enriched with data with the result that symptom-queries are barely available. You recognise these diseases in the list since they have a high %-value in the column “certainty”. This means nearly all symptom-queries regarding this disease are already answered.
 
EBMApp reliably diagnoses not only Ebola, but also Malaria, the main competitor in the differential diagnosis of Ebola. This is easy to validate if you enter the symptoms of following independent 3 confirmed case reports, which all lead to the diagnosis: 100% Malaria.
 
1. 1st case[3] 
fever, myalgia, headache, fatigue, tachycardia, dry cough, tachypnea, dyspnea, hypoxemia.
 
2. 2nd case[4] 
fever, malaise, diarrhea, icteric sclerae [=jaundice], hypotension, tachycardia, palor, sweating,  tachypnea.
 
3. 3rd case[5] 
fever, giddiness [=dizziness], chills, tachycardia, splenomegaly.
 
V. Plea
If you like EBMApp, support us with your suggestions, knowledge, manpower, funds or just start spreading the word to stop spreading Ebola. Right now we need help to collect the remaining data from different internet-sources.
 
Regards
 
Dr. Can Ansay (Social Entrepreneur)
can@ansay.de
+49-40-445589

[1] Source: http://www.nydailynews.com/life-style/health/dallas-ebola-patient-103-degree-fever-er-visit-records-article-1.1970141
[2] Source: Ergonul, Emerging Infectious Diseases (Development in Emerging and Existing Infectious Diseases), 1. edition, 2014, p. 124 ff.
[3] http://www.scielo.br/scielo.php?pid=S1413-86702005000500011&script=sci_arttext&tlng=pt
[4] http://www.elsevier.pt/en/revistas/revista-portuguesa-pneumologia-320/artigo/severe-malaria-clinical-case-9018655
[5] http://www.apjtb.com/zz/2011s2/36.pdf
 
 

0 comments

Join the conversation:

Comment