Ethical clearance for this study was obtained from the Tropical Diseases Research Centre Institution Review Board (registration number: 00002911). The questionnaire contained an information sheet regarding the study and an informed consent statement for participants to agree to participate or not. All participants who declined to take part in the study were immediately withdrawn and could not proceed to respond to the questionnaire. Access to the online portal for the questionnaire was limited to investigators who were assigned user rights by the principal investigator to ensure confidentiality of collected data.

We used the methods for calculating the sample size for prevalence studies as described by Pourhoseingholi and others.¹⁸ Assuming an 82% prevalence of good knowledge and practice regarding COVID-19 obtained from a similar study among HCWs,¹⁹ a confidence level of 95% and a precision level of 5%, we obtained a minimum required sample size of 227 participants. By extrapolating to a finite population of 1900 medical laboratory personnel officially recognised under the medical laboratory register of the Biomedical Society of Zambia, we obtained a final sample size of 204 participants.

This cross-sectional survey was conducted among 208 medical laboratory professionals in Zambia. Due to COVID-19 related restrictions imposed during this period, it was not feasible to conduct face-to-face interviews and therefore we administered an online questionnaire using Google Forms (Alphabet, Inc., Mountain View, California, United States).

Content validation of the questionnaire was done by administering the questionnaire to faculty in the Biomedical Sciences unit at the Copperbelt University, Ndola, Zambia. The results from the pre-test were not included in the final analysis but were used to modify the questionnaire based on feedback. The questionnaire had two main components: demographics and knowledge, attitudes and practices. The demographic section had 8 questions while the knowledge, attitudes and practices section was divided into the following subsections: knowledge (8 questions), attitude (5 questions) and practices (13 questions).

To ensure that only target respondents participated in the survey we distributed the link to the survey through the email database and WhatsApp (Facebook Inc, Menlo Park, California, United States) group facilitated by the Biomedical Society of Zambia during data collection from 10 to 29 June 2020. A total of 750 participants were availed the link for the questionnaire. There were 210 medical laboratory personnel who participated in the study, giving a response rate of 28%. Participants who consented to participate in the study by checking the consent box in the preliminary page of the questionnaire were allowed to proceed with the rest of the questionnaire. Only participants who completed the questionnaire by clicking on the ‘Submit’ button had their responses recorded.

Collected data were downloaded and cleaned in Microsoft Excel (Microsoft Corp., Redmond, Washington, United States) and exported to Statistical Package for Social Sciences version 23 (IBM Corp., Armonk, New York, United States) for statistical analysis. To ensure the internal consistency and reliability of the data, we used Cronbach’s alpha coefficient according to methods described in a previous study.²⁰ We obtained an alpha value of 0.645, indicating adequate reliability.²¹ The outcome variables in this study were COVID-19 knowledge levels and practice towards COVID-19. Attitude was investigated but limited to descriptive analysis due to the limited number of questions. Eight questions were used to assess COVID-19 knowledge, each scoring 1 point for a correct answer. Thirteen questions were asked to determine whether a participant had good practice towards COVID-19 or not, each scoring 1 point for positive practice and 0 for negative practice. Bloom’s cut-off of 80%²² was used to determine good knowledge and good practice. Questions were adapted from a previous study.²³

To identify factors that predict good knowledge and practice towards COVID-19, bivariate analysis was performed. All factors that were statistically significant as reported by 95% confidence interval (CI) in bivariate logistic regression were included in a forward stepwise multivariate logistic regression model to identify factors that were independently associated with good knowledge and practice.

A total of 208 medical laboratory professionals from seven provinces of Zambia took part in this study. There were more men (n = 121, 58.2%) than women. Half of the participants were aged 20–29 years while more than half had a diploma. Most of the respondents were from a hospital laboratory (n = 134, 64.4%) (Table 1).

Participants showed good knowledge regarding awareness of practical measures to stop the spread of COVID-19 (n = 208, 100.0%), the techniques used to test for COVID-19 (n = 208, 100.0%) and symptomatic management of COVID-19 (n = 202, 97.1%) (Table 2 and Table 3). One-quarter (n = 52, 25.0%) of the participants reported being involved in sampling for COVID-19 while slightly over a third (n = 73, 35.1%) reported having been trained in sample handing and transportation.

A high proportion of respondents were knowledgeable about COVID-19 (n = 175, 84.1%). Bivariate logistic regression analysis showed that having a bachelor’s degree compared to having a certificate or diploma (crude odds ratio (OR): 4.68, CI: 1.07–20.44) and COVID-19 training (crude OR: 8.72, CI: 2.02–37.65) among participants were associated with COVID-19 knowledge. Therefore, participants with higher academic qualifications and COVID-19 training were 4.68 and 8.72 times more likely to have good COVID-19 knowledge (Table 4).

Poor practices towards COVID-19 were recorded among three-quarters of the participants. Having a master’s degree or a Doctor of Philosophy compared to having a certificate or diploma (crude OR: 4.51, 95% CI: 1.30–15.70), private or research laboratories compared to clinic or health centre laboratories (crude OR: 3.09, 95% CI: 1.01–9.45) and COVID-19 training (crude OR: 12.97, 95% CI: 6.19–27.18) were associated with good COVID-19 practices (Table 5).

About 93.8% of participants reported that they would accept isolation from the community if diagnosed with COVID-19. A few (46.6%) would accept to be vaccinated against COVID-19 if a vaccine was available. On the other hand, many participants (97.6%) were ready to take part in anti-epidemic community activities. Our study revealed that 77.9% were confident that Zambia could win the battle against the COVID-19 virus (Table 6).

After controlling for possible confounding factors, having a higher current qualification (bachelor’s degree) and COVID-19 training were independently associated with good COVID-19 knowledge among biomedical professionals in Zambia. Similarly, having a higher current qualification (master’s degree or a Doctor of Philosophy) and COVID-19 training were independently associated with good COVID-19 practice (Table 7).

The study may be susceptible to self-presentation bias as it was based on an online questionnaire. The study was limited by the level of responses available and could have been strengthened by increasing the range of answers using, for example, a five-point Likert scale. The questions on attitude were limited and therefore not powerful enough to generate meaningful conclusions on the attitude of respondents. The study could not capture dropouts and respondents who refused consent as only individuals who consented and submitted the questionnaire had their responses recorded.

Our study found that medical laboratory professionals in Zambia have good knowledge regarding COVID-19. The current qualification and COVID-19 training were independently associated with COVID-19 knowledge and practice. As cases of COVID-19 continue to be recorded in the country, there is need for continuous professional development among medical laboratory personnel as a key intervention in improving their contribution to COVID-19 control efforts.