Data and sample collection using the Open Data Kit with sample processing in the field

By Kihara Absolomon

Table of Contents

Data and sample collection using the Open Data Kit with sample processing in the field

Hardware

Software

Consumables

Method

Points to be strictly followed

Hardware

1. Smart phones or tablets running on Android Operating System (OS)
2. External Bluetooth GPS Dongles
3. Netbooks
4. Barcode Scanners

Software

1. Open Data Kit suite of tools (Data collection softwares)
2. Ukasimu (Aliquoting Software)
3. Barcode generation software

Consumables

1. Labels with barcodes for sample identification
2. Barcoded Cryo vials (using generated labels)
3. Plastic boxes
4. Permanent marker pens
5. Barcoded Vacutainers (using generated labels)
6. Other sampling consumables

Method

1. Procure all the necessary hardware (as listed above).

A smart phone or tablet running on Android will suffice. The Open Data Kit (ODK) suite of tools will be used for sample collection. ODK only runs on the Android OS.

The external Bluetooth dongle is only used when conducting a rapid collection of samples from animals, where even a 1 second delay can cumulatively cause delays in the sampling exercise. While the smart phones or tablets have inbuilt GPS which is good for general use it is not suitable for rapid data collection. The built in GPS of the smart phones and tablets takes about 2-3 minutes to capture GPS coordinates and this will cause delay in the sampling process.

The barcode scanners are necessary for scanning the barcodes on the vacutainers and on the cryo vials.

2. Install either Windows or Linux OS on the netbooks. Ukasimu can be used on either platform.
3. Design the sampling form to be used and then code the form in an excel spreadsheet for later encoding to the XML format.

Designing the form is usually done in the traditional way where the team decides which kind of data needs to be collected from the field. In many instances, this is usually household level data and animal data. In addition to all the data that the scientists want to collect, a GPS coordinate must be captured at every point possible. Once the form has been encoded, it must be converted to an excel spreadsheet. This is a fairly technical process though very straightforward. Using the guidelines which can be found at: http://opendatakit.org/help/form-design/xlsform/ users can be able to transform the designed form into an excel spreadsheet.

4. Upload the excel spreadsheet to the Biorepository Server.

On uploading the excel spreadsheet to the server, the systems will verify the correctness of the spreadsheet, convert it to an xml format if all is ok and then automatically upload it to the ODK server, ready to be used in a sampling exercise.

5. Conduct a mock field exercise to test the suitability of the systems. In addition download the test data to ensure that the data collected is in an analysis friendly format.

The mock exercise is extremely important especially for first time users. It presents an opportunity to do a mock field exercise and test the systems in a near real field situation. It also helps users know what to expect at the end of the sampling exercise.

6. Generate the necessary labels for the sampling trip.

Different sample types have different label prefixes, eg. Prefix BDT for vacutainers with EDTA, BSR for plain tube vacutainers, WBL for whole blood in cryo vials, SER for serum in cryo vials, etc. All the barcodes generated are unique and are tracked by the label generation software. All vacutainers and cryo vials will bear unique random labels.

7. Prior to going to the field, affix all the generated labels to the vacutainers and the cryo vials.

This allows room for a random vacutainer to be associated to a randomly sampled animal during sampling and a random cryo vial to be associated to a random vacutainer during aliquoting.

8. While in the field select and restrain the animal to be sampled and collect the necessary samples from animal.
9. Collect the animal metadata by filling the sampling form in the phone or tablet using ODK.
10. At the end of the sampling form, associate the collected samples to the animals by scanning the samples barcodes into the animal sampling form.
11. Store the samples appropriately. Whole blood should be put in a cool place, preferably a cool box with ice packs, while serum samples should be placed in a position where the blood can easily clot at room temperature.
12. Once all the samples have been collected, transport them to the field lab.
13. In the field lab, spin the serum at the appropriate speeds and aliquot them into the appropriate number of aliquots.
14. Aliquot the whole blood into the appropriate number of aliquots.
15. Using ukasimu, associate the parent tubes and their aliquots using the method below.

1. Start ukasimu which is already installed in the netbook. This is usually by starting a browser and navigating to http://localhost/ukasimu
2. Select “Aliquot Samples”
3. Specify the parent format, aliquot format, Number of aliquots expected.
4. Click Save and then specify the box label format, purpose of the aliquots as description and the number of slots in the sample box.
5. Click Save again.
6. Place the cursor in the find field and prepare the parent samples and the aliquots for scanning into the system.
7. Scan the 1st parent sample, followed by the aliquots.
8. Place the aliquots in the box as shown by the system.
9. Correctly label the boxes as shown by the system. Place the aliquots in the slots shown by the system. Repeat the steps 15.g and 15.h above to save all the aliquots and associate them with their parents.

For each aliquot scanned after a parent sample, ukasimu interprets that that aliquot belongs to the earlier scanned parent sample.

In addition ukasimu determines the name of the box and the position to store the aliquot in. This is based on the box label format entered in the Sort Samples Home Page. This information can be edited before scanning the next sample. The action of scanning the next aliquot/parent sample will save the previous aliquot. Confirm the determined box name before scanning the next aliquots.

While saving the aliquots ukasimu ensures that:

1. The parent sample and the aliquots are of the pre-determined format.
2. The aliquots are unique and have not been saved before.
3. The determined position for the aliquots is unique and is empty.

Points to be strictly followed

1. DO NOT ENTER THE BARCODES MANUALLY! IT DEFEATS THE PURPOSE OF THE PRINTED BARCODES.
2. Label the boxes exactly as the name appears in the system.

The parent format, is the format of the parent label. eg. Whole Blood parent samples in vacutainers will have a barcode like BDT007852 or BDT078514. The barcode is composed of a prefix of 3 letters “BDT” and a suffix of 6 digits “007852” or “078514”. All vacutainers labels will follow this form of coding a set number of letters comprising the prefix and a set number of digits comprising the suffix, in this case BDT is the prefix followed by 6 digits as the suffix. The parent format is the prefix followed by a space and then the number of suffix, ie. “BDT  6”, meaning a prefix of BDT followed by 6 digits.

The aliquot format and box label format follows the same principle as the parent sample format above.