Scanwell eReader: Interpreting the RDT Results & Troubleshooting
Lisa Basel
Gillian Javetski
Results
If the image captured is good quality and there are no errors flagged by the app, one of the following result classifications may be reported:
Result | Code | Meaning | Result copy |
INDETERMINATE | -2 | The algorithm was able to detect a control line but could not confidently determine whether a test line was present or absent, possibly due to an issue with the assay (insufficient blood clearance) or an obscuring shadow. In practice, it is common to test again in this scenario. | The algorithm could not determine whether a test line was present or absent. This could be because of high background on the test strip. If you think this result is incorrect, please move to a location with more even lighting and take a picture again. |
INVALID | -1 | The algorithm was unable to detect a control line. | This test appears invalid because the control line could not be detected. If you see a control line, please ensure no shadow or glare is obscuring the control line and then take a picture again. |
NEGATIVE | 0 | The algorithm detected a control line only. | Negative |
FIRSTLINE | 1 | The algorithm detected a control line and the first test line. Depending on the RDT, this may represent P.falciparum or P.vivax. See next column. | [For SDBIO and CARESTART] P. vivax detected |
[For FIRSTRESPONSE] P. falciparum detected | |||
SECONDLINE | 2 | For RDTs with two test lines only -- the algorithm detected a control line and the second test line. For the RDTs currently supported by the library, this represents P.falciparum. | [For SDBIO and CARESTART] P. falciparum detected |
BOTHLINES | 3 | For RDTs with two test lines only -- the algorithm detected a control line and both first and second test lines. For the RDTs currently supported by the library, this represents co-infection with P.vivax and P.falciparum. | [For SDBIO and CARESTART P. vivax and P. falciparum detected |
Understanding the results
As shown in the diagram on the previous page, for valid tests (i.e. where the control line is detected), the number of possible result states depends on the RDT.
The logic the library uses for determining whether a result is invalid, valid and/or indeterminate is as follows:
If the control line is absent or uncertain:
The test is invalid.
Else: (the control line is present)
If any of the test lines is uncertain: (even if another test line is present)
The result is indeterminate.
Else if there is an obscuring anomaly:
If any of the test lines is absent:
The test is indeterminate (it could be a false negative).
Else: (all test lines are present)
The result is positive. (E.g. P.f.+P.v.)
Else: (no obscuring anomaly)
Determine the positive or negative result according to which test lines are present: negative, P.f., P.v., or P.f.+P.v.
What to do in the case of an indeterminate result
In practice, the indeterminate result is likely caused by poor background clearance (i.e. blood in the assay window) or an obscuring shadow. When an indeterminate result is returned, try the following:
Observe the assay window. If the assay window is very red or discoloured due to poor clearance of the sample (perhaps due to insufficient buffer being added or hemolysis), try again with a new test.
If the indeterminate result seems to persist but the assay window looks clear of any background colour, try moving to a different location and trying to capture the image again.
What to do in the case of an invalid result
If the app returns an invalid result, this means that a control line was not found. This could be due to:
Insufficient buffer and/or sample application leading to incomplete migration across the assay window which causes no control line to develop,
A manufacturing defect or degradation in storage which causes no control line to develop,
Poor clearance of the assay window leading to obscuring of the control line (i.e. the assay window is so red that the app cannot detect the control line),
Shadow obscuring the control line, or
Some instances of glare which may pass the glare check but obscure the control line from being detected.
If you do see a control line, please ensure no shadow or glare is obscuring the control line and then take a picture again. Otherwise, in general, you should test again using a new RDT in the case of an invalid result.
Troubleshooting
Tips to successfully take a picture:
Place the RDT on a flat surface with bright, even lighting
Place the RDT on a non-white background
Avoid having any other objects in the picture
Avoid casting shadows over the RDT when taking a picture of it
Avoid direct light reflecting off of the surface of the RDT when taking a picture of it
Take your time to line up the guide on screen with the RDT, so that the RDT fills the green rectangle as closely as possible
Try to take a clear, sharp picture of the RDT
Limitations
To date, these are the known limitations of the app. Namely, the app cannot currently:
Reliably detect the RDT on a white background
Reliably detect the RDT or the assay window when strong shadows are present
Reliably alert the user if there is motion blur that does not blur the entire image
Reliably alert the user if the image is compromised by artifacts related to low light, such as image noise and noise suppression
Reliably interpret the RDT when there is shadow across the assay window (algorithm may return “indeterminate”).
For these reasons, it is particularly important to only press “Submit for Analysis” if you have confirmed that the picture is of good quality.
Support
For questions, or to report a problem, contact the Scanwell team at hello@scanwellhealth.com with a description of the problem.
Developer Information
The Scanwell RDT Reader App is developed by:
Scanwell Health
800 Wilshire Blvd, Suite 200
Los Angeles, CA 90017
U.S.A
On this page: