Patient samples and controls
Archived, frozen samples from malaria patients with a positive diagnosis by PCR were acquired from the Provincial Laboratory for Public Health (ProvLab) in Edmonton, Alberta. The blind panel run on the Gelcycler was generated from samples collected in 2011–2012 in Uganda from consenting pregnant women participating in a study to genotype parasite resistance  (during antenatal care, at delivery and those with fever attending outpatient clinics). Blood (4 mL) was sampled by venipuncture, stored in EDTA vacutainers, and transferred to the Uganda Virus Research Institute for storage at −20°C. Thick smears were stained by Giemsa and malaria parasites counted against 200 leukocytes and expressed as number of parasites per μL of blood assuming a standard leukocyte count of 8,000/μL of blood. A smear was considered negative after examining a minimum of 100 high power fields with no parasites seen. Microscopy was performed by a laboratory technician at one of two health facilities in Mukono district. Coded samples were aliquoted and shipped frozen for testing in Alberta. Ethical approval was granted by the review boards at the Uganda Virus Research Institute and Uganda National Council for Science and Technology (Reference HS 747). Use of samples from ProvLab and the Ugandan study was approved by the Health Research Ethics Board at the University of Alberta.
The blind panel run on the Accutas comprised randomly chosen EDTA samples as positive and negative for malaria DNA that were submitted during the first seven months of 2011 for routine malaria diagnosis at the Department of Clinical Parasitology, Hospital for Tropical Diseases .
A nested PCR reaction  to amplify malarial DNA was performed on all these samples in addition to routine diagnosis by microscopy. All PCR positive samples were further characterized against the WHO International Standard for Plasmodium falciparum by a real-time PCR reaction . Once a definitive malaria diagnosis had been made for each sample the remainder were anonymized and stored as surplus to diagnostic requirements for use in ethically approved research projects. The panel consisted of positive samples infected with the following species: P. falciparum (n = 25), Plasmodium vivax (n = 2), Plasmodium ovale (n = 2), and Plasmodium malariae (n = 1). Parasitemia ranged from <5 to >1,000 parasites/μL blood. The panel also included eight negative samples.
Plasmodium knowlesi purified genomic DNA (Malaria Research and Reference Reagent Resource Center (MR4), Manassas, VA) was diluted to 2 ng/mL in non-infected blood to serve as a reference for this species. The negative control used in this study was a hemoglobin tri-level control, level 2 (Stanbio, Boerne, TX) diluted 1:10 in water. The positive control had the same composition as the negative, with the addition of a consensus PCR product to a final concentration of 105 copies/μL for hydrogel wax chips and 102 copies/μL for hydrogel plastic chips.
Plasmodium falciparum parasites were grown in human erythrocytes at 3-5% haematocrit as described . Ring stage parasites were enriched by sorbitol lysis , counted, and diluted in whole blood. The diluted parasites were frozen until use.
The hydrogel mastermix contains reagents described previously for direct amplification of DNA targets from blood  and reagents for forming a polyacrylamide gel. This mastermix was prepared with 1X Klentaq Buffer (containing 3.5 mM Mg2+; DNA Polymerase Technology, St. Louis, Mo), 40X SYBR Green I (Life Technologies, Grand Island, NY), 1X PCR Enhancer Cocktail (PEC-1) (DNA Polymerase Technology), 200 nM forward and reverse primers, 200 μM dNTPs, 0.03% BSA, 4% acrylamide (Sigma St. Louis, Mo), 0.4% bis-acrylamide aqueous solution (BioRad, Hercules, CA), 0.06% azobis (Wako, Richmond, VA), and 0.1% TEMED (Sigma). For a 100 μL mastermix volume, 3 μL of Omni KlenTaq enzyme was added (DNA Polymerase Technology). Both consensus PCR primers [15, 20], and species-specific primers [21, 22] target regions of the 18S rRNA gene of Plasmodium.
Hydrogel wax chips
A prototype consisting of a wax cassette in a 23.5 × 32 mm aluminum pan was created wherein the hydrogel is polymerized and desiccated on a coverslip above a wax-bound “trench”. A polydimethylsiloxane (PDMS) stamp was used to imprint 4 trenches in the wax. Three coverslips (20 mm × 5 mm) were placed equally spaced on top of the wax, perpendicular to the trenches, creating 12 covered slots for hydrogels. Coverslips were sealed onto the wax by briefly heating the surface of the coverslip between each trench with a soldering iron, allowing the wax underneath to melt slightly. A 13 μL aliquot of mastermix was added to each slot, creating a “gel strip”. The gel strip was photopolymerized at room temperature under UV light (367 nm) for 30 minutes. Polymerized gels were desiccated under vacuum (23 inHg) for 1 hour at room temperature and used within seven days. Chips were stored in dark conditions to prevent photo-bleaching of SYBR green dye. Coverslips were treated with 3-(trimethoxysilyl)propyl methacrylate to promote adherence to the gel strip as described . On desiccation, the gel shrinks away from the trench and stays attached to the coverslip, creating a channel underneath the strip where the sample can diffuse in. Whole blood was diluted 1:10 with water, then 8.5 μL of this sample mixture was applied underneath the coverslip/gel-strip. After the addition of samples and controls, the wax cassette was incubated for 10 minutes at room temperature then placed directly on the Peltier element within the Gelcycler for thermocycling.
Hydrogel plastic chips
Plastic chips were manufactured by miniFAB (Scoresby, Australia) by injection molding of cyclic olefin polymer (COP1420R) into top (27 × 27 mm) and bottom pieces (27 × 42 mm). To form the hydrogels, a 15 μL aliquot of mastermix was first added to the wells in the top piece. Next, the top piece was placed in a customized UV illumination chamber and flooded with nitrogen for 2 minutes. The gel was photopolymerized at room temperature under UV light (405 nm) for 30 minutes. The chamber was flooded with nitrogen a second time for 2 minutes, halfway into the UV treatment. This sustained nitrogen overlay was needed to prevent oxygen from inhibiting gel polymerization. The polymerized gels were desiccated under vacuum (23 inHg) at least 20 hours at room temperature. To complete the process of making a chip, a 10 μL aliquot of wax was added to each reservoir in the bottom piece, and the top and bottom pieces were fitted together. Wax in the bottom reservoir melts during thermocycling to cover the gel and sample ports, preventing evaporation and further cross-contamination that may occur at high denaturation temperatures. A two-sided tape adhesive on the bottom piece seals the chip upon assembly and prevents cross-contamination between gel chambers. The assembled plastic chips were used within 24 hours of preparation, or vacuum-sealed in plastic, stored at room temperature (22°C) in the dark, and tested within one month. With this design, sample added via the sample port in the top piece enters the lower reservoir and rehydrates the gel in the well. However, to ensure efficient loading of the sample in the validation experiments presented here, 8.5 μL of whole blood diluted 1:10 in water was added directly to the gel in the well before assembly of the chip. Samples were incubated prior to thermocycling as described above. As opposed to the wax chips, which sit freely on the Gelcycler Peltier element, the plastic chips were secured onto the Peltier with a gantry to ensure evenly distributed thermal transfer.
Instrumentation: Gelcycler and Accutas
The Gelcycler features a Peltier element for sample heating, components for stringent thermal regulation, a laser and CCD camera for capturing fluorescence . Consensus PCR reactions for the hydrogel wax chip were performed with a 10 minute initial denaturation step at 94°C, followed by 40 cycles of 94°C for 20 seconds, 58°C for 30 seconds, 72°C for 30 seconds, and a 2 minute final extension of 72°C. Reaction conditions for the hydrogel plastic chips were the same as above, except the consensus PCR annealing temperature was 60°C, and the species-specific annealing temperature was 64°C. MCA from 70-90°C was performed immediately after PCR. Fluorescent images were obtained and analysed as described  to generate real-time PCR and MCA curves using in-house software that requires manual analysis. A five-parameter sigmoid curve-fitting method using the first derivative was created to calculate cycle threshold (Ct) values. Samples were considered positive if they had a Ct value of less than 40 and the expected MCA peak profile. Conventional real-time PCR confirmation was performed on panel samples as described in Taylor et al. . The Accutas system refines the Gelcycler and brings many improvements. The modular design incorporates Design for Manufacturing principles to enable low cost, high volume manufacturing of the system. The Accutas system includes an LED light source rather than a laser-based system, allowing lower power and safer operation. The temperature-controlled environment allows the Accutas to be run in a wider window of temperatures. The temperature control of the Accutas was optimized and is easily adaptable to new designs. The new aluminum framed enclosure is more rugged for transport in the field. The PC control software was rewritten as well to provide an easier, step-by-step user interface with automated reporting at the end of each experiment.
Analyses were performed using SAS/STAT 9.1. Confidence intervals (95%) were calculated using the Wilson score interval method.