Medical injections, if given safely, can save lives by preventing and treating disease. Each year, at least 16 billion injections are administered in low- and middle-income countries, but over half of these injections have been shown to be either unnecessary or unsafe.1–4 Unsafe injections can cause bacterial infections leading to abscesses and septicemia.1 Additionally, they are important routes of transmission of blood-borne pathogens.2 These pathogens include hepatitis B virus, hepatitis C virus, HIV, malaria, trypanosomiases, and viral hemorrhagic fever viruses. In 2000, it was estimated that unsafe medical injections led to 260,000 HIV infections, 2 million hepatitis C virus infections, and 21 million hepatitis B virus infections globally.5
Poor disposal of contaminated needles and syringes and other medical waste poses a risk to health care providers, patients, and the community.6 Traditionally, most health programs have concentrated their efforts on disposal of syringes and needles within the health care settings.7 With expansion of health services and overwhelmed health facility waste management systems, there is increased disposal of medical waste, including contaminated needles and syringes, in the community.6–8 This has led to increases in risk of injury to community members, especially children and domestic waste handlers.9 Whereas industrialized countries such as the United States of America, Canada, and Australia have community-based programs for needle and syringe disposal,10 this is not the case in most low- and middle-income countries, including Kenya.
The Kenya Ministry of Health recognizes that administration of unsafe and unnecessary injections and inappropriate disposal of the medical waste are health challenges.11 In 2008, it was estimated that 2.2% of new HIV infections in Kenya were a result of unsafe injections received in health facilities.12 The first Kenya AIDS Indicator Survey (KAIS 2007) reported that 33% of adults and adolescents aged 15–64 years had received at least 1 injection within the 12 months preceding the survey and that about 50% preferred injections over oral medications.13 Furthermore, there was a corresponding increase in HIV prevalence with increasing number of medical injections reported.
Since 2004, the Ministry of Health has implemented several strategies to address injection safety, including health care provider trainings, behavior change communication, and community education on the dangers of unsafe and unnecessary injections.11 It has also educated the public on its role in ensuring safe injections and preventing re-use among patients by confirming that the injection device used is new.14 However, many people remain unaware of the increased health risks associated with unsafe administration of injections.14
Despite the ongoing interventions, no population-based data on injection preference, injection usage, and medical waste disposal in the community have been collected since KAIS 2007. The second Kenya AIDS Indicator Survey (KAIS 2012) presented an opportunity to obtain current information on medical injection safety and medical waste disposal, allowing for an assessment of trends over the past 5 years.
Study Design and Study Population
KAIS 2012 was a stratified 2-stage cluster population-based survey designed to generate national and sub-national estimates of HIV-related indicators among Kenyans aged 18 months to 64 years. The survey methods for KAIS 2012 are described in detail elsewhere.15 Briefly, individual questionnaires were administered to adults and adolescents aged 15–64 years that collected information on sociodemographic characteristics, behaviors, and access to health care services, including medical injection practices, injection safety, and medical waste disposal. Participants were asked to provide blood samples for HIV testing at a central laboratory. Participants were informed that their HIV test results from the laboratory would not be returned to them. However, they were offered home-based HIV testing and counseling to learn their HIV status in the home using rapid HIV test kits according to the Kenyan national HIV testing and counseling protocol.15 In this article, we present a subanalysis of KAIS 2012 to describe medical injection use, medication preferences, and reports of unsafe medical waste disposal in the community among Kenyan adults and adolescents aged 15–64 years.
We had three outcome variables of interest: history of medical injection in the past 12 months, medication preference, and HIV infection. Previous history of medical injection was based on the question, “Have you had an injection for any reason in the last 12 months?” For female participants, we also based history of medical injection on the question, “Are you currently using any method to delay or avoid getting pregnant?” Those who answered “Yes-injectable method” and had not been captured in the question above were classified as having a history of medical injection in the past 12 months. Predictor variables for history of medical injection in the past 12 months included sex, age, marital status, level of education, area of residence, region, household wealth index, and medication preference. Medication preference was based on the question: “If you had a choice, would you like to receive medication as an injection or a pill?” Predictor variables for medication preference included demographic variables and history of medical injection in the past 12 months, number of injections, and use of injectable family planning method. HIV status was based on the laboratory-confirmed HIV results at the central laboratory. The predictor variables for HIV infection were age, injection history, use of injectable contraceptives (for women only), the number of injections received in the past 12 months, and whether the injection was self-administered.
Participants were requested to provide a blood specimen for HIV testing at the National HIV Reference Laboratory in Nairobi. Specimens were tested for HIV antibody using the Vironostika HIV-1/2 UNIF II Plus O Enzyme Immunoassay (bioMérieux, Marcy d'Etoile, France), and positive results were confirmed using the Murex HIV.1.2.O HIV Enzyme Immunoassay (DiaSorin, SpA, Saluggia, Italy). Repeat testing was performed for discordant results using the same algorithm, and if results remained discordant, final results were obtained using polymerase chain reaction (Cobas Amplicor HIV-1 Monitor Test, version 1.5, Roche Molecular Diagnostics, Pleasanton, CA).
Data Management and Analysis
Data were collected at the point of interview using netbook computers (Mirus Innovations, Mississauga, Ontario, Canada). Data collected in the household were transmitted through a secure wireless network to a central database in Nairobi where data merging, cleaning, and weighting were performed before analysis. Statistical analyses were performed using survey procedures in STATA version 12.0 (STATA Corporation, College Station, TX) and SAS version 9.3 (SAS Institute Inc., Cary, NC). We conducted descriptive analyses to describe persons who had received a medical injection in the past 12 months by select demographic and injection-specific characteristics and present results as weighted population proportions and 95% confidence intervals (CI). We conducted bivariate analyses to estimate the frequencies and proportions of receiving injections in the past 12 months, medication preference, and HIV infection by select variables. Using variables that had a P value <0.1 in the bivariate anlayses, we also conducted multivariable analyses to determine factors independently associated with receiving injections in the past 12 months and medication preference. Estimates for measures of association are presented as odds ratios, adjusted odds ratios (aOR), and 95% CI. All analyses were weighted to account for sampling probability and to adjust for survey nonresponse.
The survey protocol, survey collection tools, and consent procedures were approved by the Kenya Medical Research Institute (KEMRI) Ethical Review Committee, the US Centers for Disease Control and Prevention (CDC) Institutional Review Board, and the Committee on Human Research of the University of California, San Francisco (UCSF).
A total of 13,720 adults and adolescents aged 15–64 years participated in the survey. Of these, 13,673 (99.7%) completed the question on having received or not received an medical injection for any reason in the previous 12 months; 7928 (51.0%) of these were women, and 5745 (49.0%) were men.
History of Medical Injection in the Previous 12 Months
In total, 4906 (35.9%, 95% CI: 34.5 to 37.3) had received at least 1 medical injection in the previous 12 months. Among those, 63.6% (95% CI: 61.9 to 65.4) were women (Table 1). Over 60% were married or cohabiting (64.2%, 95% CI: 62.5 to 65.9), lived in rural residences (62.1%, 95% CI: 59.5 to 64.6), and aged less than 35 years, including 32.0% (95% CI: 30.5 to 33.4) of those aged 15–24 years and 32.8% (95% CI: 31.3 to 34.2) of those aged 25–34 years. Injection recipients were evenly distributed by household wealth index with 17.3% (95% CI: 15.0 to 19.7) in the poorest and 20.3% (95% CI: 17.6 to 23.0) in the richest wealth index quintile.
TABLE 1-a Select Cha...Image Tools
Among respondents who reported receiving at least 1 injection in the previous 12 months, 96.3% (95% CI: 95.5 to 97.1) were administered by health care providers, 4.6% (95% CI: 3.5 to 5.7) were self-administered, and 0.4% (95% CI: 0.2 to 0.7) were administered by traditional practitioners. Injectable contraceptives accounted for 42.2% (95% CI: 40.0 to 44.4) of all medical injections administered to women (data not shown). The vast majority of respondents (95.9%, 95% CI: 95.2 to 96.7) reported that the last time they received an injection from a health care provider, the needle came from a new, unopened package. Half (50.2%, 95% CI: 48.4 to 52.0) preferred injections as medication, 36.5% (95% CI: 34.8 to 38.3) preferred pills, and 13.3% (95% CI: 12.1 to 14.4) had no treatment preference. Seven percent of injection recipients (7.4%, 95% CI: 6.4 to 8.4) had seen a used needle or syringe near their home or community in the past 12 months compared with 6.3% (95% CI: 5.5 to 7.0) of all survey respondents. In total, 14,982 injections received by participants were administered by health care providers in the previous 12 months, corresponding to an average of 3.6 injections per person among those that received at least 1 injection in the past year. When applied to the survey population, the per capita medical injection rate for adults and adolescents aged 15–64 years was 1.1 injections per person per year.
TABLE 1-b Select Cha...Image Tools
In multivariate analyses, factors independently associated with increased odds of receiving a medical injection in the previous 12 months were being a woman (aOR, 2.1; 95% CI: 1.8 to 2.3); being currently or previously married or cohabiting compared with being never married or cohabiting [married or cohabiting (aOR, 2.2; 95% CI: 1.9 to 2.5); ever widowed (aOR, 1.7; 95% CI: 1.4 to 2.2); and separated/divorced (aOR, 1.7; 95% CI: 1.3 to 2.2)]; living in Nyanza (aOR, 1.6; 95% CI: 1.2 to 2.0) and Western (aOR, 1.4; 95% CI: 1.1 to 1.8) regions compared with Nairobi region; increasing education levels compared with having no primary education [incomplete primary (aOR, 1.7; 95% CI: 1.2 to 2.2); complete primary (aOR, 1.5; 95% CI: 1.2 to 1.9); and secondary or higher level of education (aOR, 1.4; 95% CI: 1.1 to 1.8)]; and being in the middle (aOR, 1.3; 95% CI: 1.1 to 1.6) or fourth highest (aOR, 1.3; 95% CI: 1.1 to 1.6) wealth quintile compared with being in the poorest wealth quintile (Table 2). Compared with persons aged 15–24 years, persons aged 35–64 years had significantly lower odds of receiving an injection in the past 12 months [aged 35–44 years (aOR, 0.7; 95% CI: 0.6 to 0.8), aged 45–54 years (aOR, 0.6; 95% CI: 0.5 to 0.7), and aged 55–64 years (aOR, 0.5; 95% CI: 0.4 to 0.7)]. We also examined these associations separately for men and women and found similar results (data not shown).
Overall, 51.2% (95% CI: 49.7 to 52.8) of survey participants preferred an injection to a pill as medication (Table 3). Preference for injection was higher among women (56.7%, 95% CI: 54.9 to 58.5) than men (45.3%, 95% CI: 43.4 to 47.2). Preference for injection was highest among persons aged 25–34 years, at 56.2% (95% CI: 54.0 to 58.4) and decreased with increasing age to a low of 45.4% (95% CI: 41.4 to 49.4) for persons aged 55–64 years. Over half of those who preferred injections resided in rural areas (51.6%, 95% CI: 49.8 to 53.4), were married or cohabiting (54.6%, 95% CI: 52.9 to 56.3), and had completed primary education (52.2%, 95% CI: 50.1 to 54.3). Preference for injection was higher among those who had received at least 1 injection in the previous 12 months (57.9%, 95% CI: 55.9 to 59.8) compared with those who had received no injection during the same time period (47.3%, 95% CI: 45.6 to 49.0).
In multivariate analyses, older respondents aged 35–44 years (aOR, 0.8; 95% CI: 0.7 to 0.9), 45–54 years (aOR, 0.7, 95% CI: 0.6 to 0.8), and 55–64 years (aOR, 0.6, 95% CI: 0.5 to 0.7) had lower odds of preferring an injection to a pill for medication than younger respondents aged 15–24 years. Female sex (aOR, 1.5; 95% CI: 1.3 to 1.6), residing in urban areas (aOR, 1.2; 95% CI: 1.1 to 1.4), and receiving an injection in the previous 12 months (aOR, 1.4; 95% CI: 1.3 to 1.5) were associated with a higher odds of preferring an injection to a pill.
Medication Injection and Associations With HIV Infection
HIV prevalence was 6.3% (95% CI: 5.1 to 7.4) among individuals who received at least 1 injection in the past 12 months compared with 5.3% (95% CI: 4.6 to 5.9) among those who had not (data not shown). Among those who had received an injection from a traditional practitioner in the past 12 months, HIV prevalence was 20.6% (95% CI: 0 to 44.4). In contrast, HIV prevalence among persons who had self-administered an injection in the past 12 months was 3.5% (95% CI: 0.4 to 6.6).
Among men, HIV prevalence among those who received a medical injection in the past 12 months (5.8%, 95% CI: 4.2 to 7.3) was significantly higher than those who did not receive an injection (3.8%, 95% CI: 3.1 to 4.5) (Table 4). Men who had self-administered an injection in the past 12 months had low HIV prevalence at 2.6% (95% CI: 0 to 6.3). Although men who reported receiving 2 to 3 (6.0%, 95% CI: 3.6 to 8.3) and 4 or more (6.6%, 95% CI: 3.0 to 10.2) medical injections from a health care provider in the past 12 months had higher HIV prevalence than men who had received no injections from health care providers (3.9%, 95% CI: 3.2 to 4.6), these differences were not statistically significant. After adjusting for age and number of injections received in the previous 12 months, men who had received a medical injection in the previous 12 months were 3 times more likely to be HIV infected than men who did not receive injections in the same time period (aOR, 3.2; 95% CI: 1.2 to 8.9).
Among women, HIV prevalence was 7.1% (95% CI: 6.2 to 8.1) among those who had not received a medical injection in the past 12 months, 5.5% (95% CI: 4.1 to 7.0) among women who received injectable contraceptives, and 7.4% (95% CI: 5.8 to 8.9) among women who received medical injections for reasons other than family planning (Table 5). Women who self-administered a medical injection in the past 12 months had an HIV prevalence of 4.8% (95% CI: 0 to 10.0). No associations were observed between the number of injections received by a health care provider and HIV infection among women. After controlling for age and the number of injections administered by a health care worker in the past 12 months, women who received injections for reasons other than family planning purposes were 3 times more likely to be HIV infected compared with women who received no injections in the past 12 months (aOR, 2.6; 95% CI: 1.2 to 5.5).
In this nationally representative survey, we found that approximately one-third of adults aged 15–64 years had received a medical injection in the 12 months preceding the survey, and the vast majority of these had received injections from a health care provider. Those who received injections were mainly women, currently or previously married, with higher wealth and education, and residents of Nyanza and Western regions. The estimated number of injections from health care providers per person per year was 1.1 injections. A global average of 3.4 medical injections per person per year among children and adults has been reported elsewhere, with Africa reporting an average of 2.2 medical injections per person per year.1 The proportion of medical injection recipients in our study, however, remained similar to what was reported in KAIS 2007, where 1 in every 3 persons aged 15–64 years received at least 1 medical injection in the previous 12 months.13
We found that 1 in 15 participants had seen a used syringe or needle near their home or in their community in the past 12 months. This finding corroborates 2 recent studies reporting improper medical waste disposal in the community. Mazrui and colleagues16 found that 11% of private health facilities in Nairobi disposed of medical waste in open dump sites, and a United Nations Environmental Program study team reported seeing used needles in the general waste disposal site in Dandora, Nairobi.17
Given a choice, half of Kenyan adults and adolescents preferred an injection to a pill for medication purposes, similar to what was reported in KAIS 2007.13 Furthermore, this rate was also similar to findings from a program evaluation of safe medical injections in Kenya, which showed that 5 out of 10 persons in Western and Nyanza regions preferred medical injections to pills.18 We also found that individuals who received an injection in the previous 12 months had significantly higher odds of preferring medical injections to pills compared with those who had no medication preference. These findings indicate that injection preference can influence injection use which could potentially lead to unnecessary and unsafe injections. Women were more likely than men to prefer injections over pills and accounted for 70% of medical injections that were administered by health care providers, with at least 40% for contraceptive purposes. This was consistent with findings from the Kenya Demographic and Health Survey of 2008–09, which reported that injectable contraceptives were the most widely used family planning method, with 1 in 5 women aged 15–49 years reporting that they were currently using this method.19
Individuals who had received higher number of injections from a health care provider in the previous 12 months had similar odds of HIV infection compared with individuals who had received no injections from a health care provider in the same time frame. Both men and women who received medical injections in the 12 months preceding the survey (not for contraceptive purposes among women) were more likely to have been HIV infected than those who had not received any injections in the previous 12 months. High HIV prevalence was noted among persons who had received injections from traditional practitioners, though the number reporting this practice was small. There is a possibility of re-use of injection devices in traditional medicine settings, potentially increasing the risk of HIV transmission and acquisition. We were encouraged to find that over 95% of those who received injections from health care providers observed a new, unopened needle package being opened. Patient-observed sterile treatment is a national strategy that has been adopted in Kenya since 2004 to ensure sterile care and reduction in HIV transmission risk in health care settings.20
This study had several limitations. The analysis relied on self-reported data that may have been limited by recall bias. To minimize this bias, we restricted the recall period to the preceding 12 months from the survey. Second, variables on medical waste in the community were dependent on correct knowledge of a used syringe or needle, but no informational material was provided to survey participants to confirm visual understanding of medical waste. Therefore, our reported estimates on medical waste could be either an overestimate or underestimate of true values in the population. Because KAIS 2012 was a cross-sectional survey where potential predictors and outcomes were measured at the same time, we were not able to determine causality in associations observed, such as those reported for injection status and HIV infection. In addition, we did not collect information on whether medical injections received were curative or preventative, both of which may have impacted our estimates of injection history and injection preference. Finally, the number of self-administered injections and number of injections received from traditional practitioners were not quantified, which may have contributed to lower injection totals.
Despite these limitations, this study provides important nationally representative population-based data that can be used to inform the national program on targeted strategies for the prevention of medical transmission of HIV and other blood-borne pathogens. Efforts to improve health communication, particularly to those who are more likely to receive medical injections, are needed to reduce the risks of unsafe injections, to address the hazards of medical waste disposal in the community, and to educate traditional practitioners on safe injection practices. In addition, medical waste management programs need to support waste disposal at the health facility and the community. Finally, given the wide use of injectable contraceptives among women, we recommend that the national reproductive health program integrate injection safety practices in reproductive health services. These recommendations can form the basis for policy makers to support injection safety interventions for patients, health care providers, and the community as part of comprehensive infection prevention and control programs.
The authors thank the KAIS 2012 field teams and all the individuals who participated in this national survey. They would like to thank George Rutherford, Kevin DeCock, Amanda Viitenen, and Anthony Waruru for reviewing and providing input on the article, and the KAIS Study Group for their contribution to the design of the survey and collection of the data set: Willis Akhwale, Sehin Birhanu, John Bore, Angela Broad, Robert Buluma, Thomas Gachuki, Jennifer Galbraith, Anthony Gichangi, Beth Gikonyo, Margaret Gitau, Joshua Gitonga, Mike Grasso, Malayah Harper, Andrew Imbwaga, Muthoni Junghae, Mutua Kakinyi, Samuel Mwangi Kamiru, Nicholas Owenje Kandege, Lucy Kanyara, Yasuyo Kawamura, Timothy Kellogg, George Kichamu, Andrea Kim, Lucy Kimondo, Davies Kimanga, Elija Kinyanjui, Stephen Kipkerich, Danson Kimutai Koske, Boniface O. K'Oyugi, Veronica Lee, Serenita Lewis, William Maina, Ernest Makokha, Agneta Mbithi, Joy Mirjahangir, Ibrahim Mohamed, Rex Mpazanje, Silas Mulwa, Nicolas Muraguri, Patrick Murithi, Lilly Muthoni, James Muttunga, Jane Mwangi, Mary Mwangi, Sophie Mwanyumba, Francis Ndichu, Anne Ng'ang'a, James Ng'ang'a, John Gitahi Ng'ang'a, Lucy Ng'ang'a, Carol Ngare, Bernadette Ng'eno, Inviolata Njeri, David Njogu, Bernard Obasi, Macdonald Obudho, Edwin Ochieng, Linus Odawo, James Odek, Jacob Odhiambo, Caleb Ogada, Samuel Ogola, David Ojakaa, James Kwach Ojwang, George Okumu, Patricia Oluoch, Tom Oluoch, Kenneth Ochieng Omondi, Osborn Otieno, Yakubu Owolabi, Bharat Parekh, George Rutherford, Sandra Schwarcz, Shahnaaz Sharrif, Victor Ssempijja, Lydia Tabuke, Yuko Takenaka, Mamo Umuro, Brian Eugene Wakhutu, Wanjiru Waruiru, Celia Wandera, John Wanyungu, Anthony Waruru, Paul Waweru, Larry Westerman, and Kelly Winter.