The study was carried out in the Faranah Regional Hospital (FRH), in Guinea, a partner hospital of the Robert Koch Institute, Berlin, Germany. The FRH is a governmental reference hospital for a population of 300,000 inhabitants with an adult literacy rate of approximately 32.0% . The hospital employs 91 healthcare and administrative staff-members and is comprised of 16 wards, including surgery, laboratory and, in aftermath of the Ebola Virus Disease epidemic from 2013 to 2016, an isolation ward.
The study was conducted as part of the PASQUALE (Partnership to Improve Patient Safety and Quality of Care) project and funded by the GIZ ESTHER Alliance (Ensemble pour une Solidarité Thérapeutique Hospitalière en Réseau). The PASQUALE project responds to the first WHO Global Patient Safety Challenge: “Clean Care is Safer Care” . Ethics approval was obtained from the Comité National d’Ethique pour la Recherche en Santé, Guinea.
The study was conducted to assess feasibility and effectiveness of the WHO HH improvement strategy in this low-resource setting. All wards and currently employed HCWs (not including administrative or cleaning staff) were invited to participate. Activities consisted of four phases: preparatory phase, pre-intervention evaluation, intervention and post-intervention evaluation.
Phase 0: preparatory phase: In December 2017, a needs assessment was conducted together with the FRH staff and leadership. In this participatory assessment, it was decided to focus on HH, water supply and sterilization. Moreover, a qualitative research study on HH was conducted to gain further insight, the results of which are planned to be published separately.
Phase I: pre-intervention evaluation: In January 2018, a baseline assessment took place inviting all currently employed HCWs to participate. This baseline assessment included the surveys and questionnaires on ward infrastructure, HCWs’ perception on HH, HH knowledge, and an observation of HH practices, using the validated WHO tools . HH practices were assessed using the WHO “My 5 Moments for Hand Hygiene” approach including the indication 1) before touching a patient, 2) before clean/aseptic procedures, 3) after body fluid exposure risk, 4) after touching a patient and 5) after touching patient surroundings . The knowledge questionnaire included questions such as “What is the most frequent source of germs responsible for healthcare-associated infections”, “Which of the following HH actions prevents transmission of germs to the patient”, “What is the minimal time needed for ABHR to kill most germs on your hands” and “Which type of HH method is required in the following situations (…)”. The perception questionnaires focused on the five core elements of the WHO HH strategy (system change, education, observation and feedback, reminders in the workplace and patient safety climate) with questions like “What is the effectiveness of HH in preventing healthcare-associated infection” and “(…) how effective would the following actions be to improve HH permanently in your institution (…)”. When participants required clarification of language or help with Likert scales, support was available by members of the research team. Each participant received an identification number that was planned to be used instead of their names for pairing baseline and follow-up data on HH knowledge and perception. The direct observations were carried out by trained researchers from the PASQUALE project during day shifts at random times, without prior announcement. HH indications and opportunities were recorded throughout the observation. A priority rule was applied to ensure that only one indication was associated with each opportunity. This rule specified a hierarchy for simultaneously occurring indications as follows: before aseptic/clean procedure > after body fluid exposure risk > after touching a patient > before touching a patient > after touching the patient surrounding .
Phase II: intervention: December 2018, a tailored workshop was conducted. This training was adapted to potential improvement points of HH knowledge and practice identified in the pre-intervention evaluation during phase I. The training was held on three occasions as a one day workshop for 24 participants at a time. As such, every HCW had the opportunity to participate without interrupting routine hospital functions.
Local production of ABHR was reintroduced in a designated manufacturing room of the hospital pharmacy with four batches (10L each) per month. This production schedule was based on a national estimate for hospital-wide ABHR needs . Production of ABHR was initially introduced by WHO in 2016, but manufacturing was not sustainable due to supply issues such as access to hydrogen peroxide and peroxide test strips. These challenges were overcome by fostering collaborations with experienced Nongovernmental Organizations, such as Expertise France, and local private suppliers. The ABHR was subsequently produced following “Formulation 1” from WHO guidelines. This formulation specifies usage of ethanol 96%, hydrogen peroxide 3%, glycerol 98% and boiled, cold water [19, 22]. To apply the learnt hygiene measures every HCW received a pocket bottle of 100 ml ABHR and every ward or consultation room a bottle of 500 ml ABHR. Bottle labeling included the instruction “apply a palmful (3ml), cover all surfaces of the hands, rub hands until dry”. Upon request of the partner in Faranah who wished to be reassured about the efficacy of the locally produced ABHR, a partial efficacy testing was carried out in laboratories of the division for Hospital Hygiene, Infection Prevention and Control at the Robert Koch Institute. The efficacy testing was performed as suspension test according to the European Norm DIN EN 13727 using Enterococcus hirae as test organism. Further support for the local pharmacy is ongoing and south-south information exchange is being fostered between the FRH and the other PASQUALE Project partner Centre Hospitalier Universitaire in Bouaké, Côte d’Ivoire. A conjunct ABHR production training was held in Bouaké with assistance of the FRH pharmacists and PASQUALE team in June 2019.
As part of the participatory approach a local coordinator was given the responsibility of fostering project work within the FRH by conducting observations, regular HH reminding sessions during staff meetings and promotions of the locally produced ABHR.
Phase III: post-intervention evaluation: During December 2018 to March 2019, the assessment was repeated following the same methodology as in the pre-intervention evaluation in Phase I, with the perception survey containing one additional post-intervention part asking about perceived effects of the intervention.
Furthermore, production of ABHR was monitored and consumption of ABHR was tracked six months before (July to December 2018) and after the intervention (January to June 2019).
All data was entered in WHO preprogrammed Epi Info data templates and analyzed using Stata 15.2 (StataCorp LLC, College Station, Texas, USA). For HH knowledge questionnaire responses, a score was calculated equaling the number of correct answers (maximum score 25 points). The scores were summarized as medians and interquartile ranges. Since pairing was only possible for half of the study population (30/62), a sensitivity analysis comparing paired and unpaired Wilcoxon rank-sum tests was performed. As conclusion of both tests was the same (data not shown), only results of the unpaired Wilcoxon rank-sum test are presented in this paper. Two-tailed p-values less than 0.05 were considered to be statistically significant.
HH perceptions on the five components of the WHO HH Strategy were assessed in baseline and follow-up questionnaires. Additional post-intervention perception questions were reported as the total number and percentage of follow-up respondents answering “seven” on a seven-point Likert scale, with one equaling “not effective” and seven “very effective”.
HH compliance was calculated as the number of HH actions performed divided by the number of all opportunities requiring HH according the WHO 5 Moments of HH. Compliance at baseline and follow-up was compared using c2 tests, by wards and by professional categories. Multiple linear regression was performed to assess the association between the intervention and knowledge score, exploring the confounding effect of gender, age group, profession and ward. Multivariable logistic regression was performed with pre/post-intervention period as the main independent variable and compliance as the primary outcome. Confounders proposed in the literature “type of ward”, “hand hygiene indication” and “professional category” were included in the initial logistic regression model and maintained there if the crude OR differed substantially from the adjusted one. Consequently, the confounders “hand hygiene indication” and “professional category” were included in the final model.
As most healthcare professionals had more than one HH opportunity, the observations were not independent. For confidentiality purposes, and following the WHO multimodal HH Improvement Strategy observation form, HCWs were not identified during observation. To account for this lack of independence a design effect of two was assumed and accounted for by doubling the standard error (22); this approach has been used before in a similar study [11, 23].