The impact of occupational splash,sharps and needlestick injuries (SSNIs) on the quality of life of healthcare workers in a Kenyan university hospital

2.1 Epidemiology of SSNI and HCW risk

SSNIs are a prevalent occupational hazard amongst HCWs [1, 10]. A recent cross-sectional study in Kenya examined 1,665 cases of SSNI’s over four years from 2011–2014 and suggested incidence was increasing [11]. Ugandan research [12] (Uganda shares a border with Kenya), found an SSNI prevalence rate of 46% amongst HCWs. However, underreporting of these accidents means this may not be an accurate picture. Makhoka [13], estimated a non-reporting rate of about 40% in the hospital involved in this study. Studies elsewhere in the world show that 35% to 62% of SSNIs are not reported [14–17]. This is important to HCWs in this study because they are at greater risk because of the high prevalence of BBPs in the general population in sub-Saharan African countries.

Globally, the United Nations Programme on HIV/AIDS (UNAIDS) in 2015 stated that 36.7 million people were living with HIV, 19 million, roughly 52% of those living with HIV were in eastern and southern Africa [8]. Locally, the Kenya Aids Indicator Survey of 2012 [18] showed that the HIV prevalence rate stood at 5.6% amongst adults between 15–64 years (approximately 1,192,000 people). By comparison, there were only 98,400 people living with HIV in the UK in 2012 [19].

Besides, the HIV burden, the African continent also has high numbers of people infected with chronic hepatitis B, C and A, which puts the health of HCWs at further risk [20]. In a recent systematic and pooled data review from 1965 to 2013; Schweitzer, Horn, Krause et al. [21] estimated global hepatitis B surface antigen (HBsAg) seroprevalence for chronic hepatitis infection and reported that it was highly endemic in the WHO African Region (total 8.83%, CI 8.82–8.83). Kenya has a rate of 5.16% (CI 4.86–5.48) compared to developed countries like the USA where the rate was only 0.27% (CI 0.24–0.30).

At even greater risk, are HCWs who sustain SSNIs from source patients who are already infected with BBPs. The recent cross-sectional study in Kenya [11] revealed that out of 851 cases where source patient HIV status was known, 39.4% proved HIV positive and although only 45 (5%) of source patients status was known for Hepatitis B, 15 affected workers (36.6%) proved positive for the antigen representing infection. It additionally indicated that other risks to Kenyan HCWs included low Hepatitis B vaccination coverage, a lack of testing for antibody titres, ineffective management of post exposure prophylaxis (PEP) services, inadequate staff knowledge of risk and a lack of psychological support in the event of exposure [11].

The merging of all these factors not only creates a risky situation but is a cause of worry for HCWs who sustain SSNI’s. There are also additional PEP treatment side-effects that affected HCWs may have to bear. All these issues may impact on and influence how HCWs appraise their general well-being and life quality following exposure.

2.2 Quality of life (QOL) concept and SSNIs

The World Health Organization (WHO) described QOL as “…an individual’s perception of their position in life in the context of the culture and value systems in which they live and in relation to their goals, expectations, standards and concerns.” [22 p.1] It encompasses one’s state of physical health, psychological state, level of independence, social relationships, personal beliefs and is affected by the environment one interacts with [22]. QOL is a multidimensional concept that includes the subjective evaluation of both the positive and negative aspects of one’s life [23]. QOL is an important measure for clinical outcomes which has been applied to many chronic conditions to understand the impact of disease on people’s lives overall and looks beyond their symptoms [24].

While reports on how cancers, diabetes and other chronic conditions, impact on patients QOL, little is known about SSNIs and their impact on QOL. A few studies have explored this but their results are mainly limited to physical and psychological symptoms and not overall QOL as conceptualized by WHO [22]. For example, Zhang and Yu [25] in their Chinese study showed that 15.2% of their participants experienced emotional distress such as anxiety, worry, frustration, panic; post-SSNI, while bigger numbers, (57.6%), feared being infected from the exposure. Similarly, Gershon et al. [4] found that 85% of HCWs who were started on PEP treatment had adverse side-effects which mostly included nausea, stomach ache, fatigue, headache and diarrhea. They also found that the psychological symptoms suffered included, 53% of respondents feeling anxiety, 18% experiencing insomnia, 13% depression, and 10% experiencing a loss of appetite, another 10% affected with sleepiness, and another 10% are frequently crying. Other studies have also associated mental health problems such as depression and post-traumatic disease syndrome (PTDS) with SSNIs [7, 27]. Stigmatization and social isolation in relation to occupational exposures to BBPs through SSNIs have also been reported [5, 28–30]. Other difficulties involving both sexual and relational problems have also been highlighted as consequences of SSNIs exposure [4, 31].

QOL measurement is often considered only a health-related outcome [32]. Clinical Outcomes Models and Health-Related Quality of Life (HRQOL) are examples of some measures that have been applied in quality of life studies, but they have limitations and weaknesses since they focus mainly on health, function or symptoms related factors in what they assess and so do not comprehensively address QOL [32–35]. Gill and Feinstein [36] contend that QOL measurement can only be achieved by incorporating the person’s view of what is uniquely important to them, with regard to both their health status and their wider lives. According to the developers, Burckhardt and Anderson’s [33] Quality of Life Scale (QOLS) tool does this and has proved to be appropriate in measuring the phenomenon of QOL. It has cross-cultural applicability and is capable of measuring quality of life conceptually distinct from health status or other causal QOL indicators [33]. The Burckhardt and Anderson 16-items self-reported QOLS was originally developed by Flanagan [37] but was later refined by Burckhardt and Anderson to its current form [33].

The importance of undertaking QOL studies is that they can inform healthcare service development and efforts to seek care outcome improvement. For example, DiSipio, Hayes, et al. [38] found out that the health-related quality of life (HRQOL) of younger breast cancer survivors (<50 years of age) at 6 months was lower at a mean of 80.2 compared to that of older women (>50 years) with a mean of 87.1. Having discovered that younger women’s physical and emotional wellbeing was impaired they suggested a targeted intervention to improve younger women’s’ views of their wellbeing. Other researchers have also studied different patient groups including those with sickle cell disease and fibromyalgia [38, 39] to understand how these diseases impact on people’s QOL and what developments in their care could lead to improvement.

4.1 Data collection tools

Data collection was carried out between May to July 2013 using Burckhardt and Anderson’s [33] 16-items self-reported Quality of Life Scale (QOLS) and a 10- item self-reported symptoms questionnaire which also explored the demographic variables considered useful for analysis. The 16-QOLS items cover six broad domains; material and physical well-being, relationships with other people, social, community and civic duties, personal development and fulfilment, recreation and finally independence. Each of the 16-QOLS questions has a 7 point Likert scale where; 1 is described as ‘terrible experiencing of life’ and 7 is ‘delighted’. The total maximum possible score for QOLS is 112 points and the lowest is 16 [33]. The higher the scores, the better the person’s QOL, lower scores predict poorer QOL [33]. It is estimated that a score of 90, represents the expected average score for a general population [33]. QOLS has a high reliability and internal consistency with a Cronbach’s α between 0.88 to 0.92 noted in different studies [40, 41] making the QOLS both a valid and reliable tool for studies of this type, hence its selection for use here.

The symptoms questionnaire explored whether lack of appetite, tiredness, nausea and vomiting, difficulty sleeping, anxiety and worrying, reduced the desire for sexual intimacy, nervousness and despairing about the future, feeling depressed and lastly diarrhea were experienced post-SSNI and during PEP treatment. The symptoms questions and responses were also presented as Likert scales from 1 to 6; with score 1; being ‘symptoms experienced all the time, more than once a day’. A score of 6 meant ‘no symptoms were experienced at all’. The lower the scores the worse the impact of experienced symptoms would be and vice versa.

The questionnaire also included one open-ended question to collect some views from the respondents on their SSNIs experience.

6.1 Socio-demographic characteristics of participants

After data cleaning, 416 completed responses were received from the 1005 eligible participants giving a response rate of 41.4%. This was similar to the response rate of 46% achieved in a similar hospital-wide study of SSNI’s in Uganda [12]. Of the 416; 55.3% were females and 44.7% male. The age bracket 25–34 years was the most populous which is reflective of wider Kenyan demographics [11]. Many of the respondents (25.7%) had at least a diploma level of education. Nurses, more than any other cadre of staff were the most likely to respond (25.4% of the total). Amongst the participants; 46.2% (n = 192) answered: “Yes”; while 53.8% (n = 224) responded “No” to having had experience of SSNIs in the previous five years (Table 1).

Of the 192, “yes” responders (Table 2), the majority were within the age bracket 35–44 years (39.1%), followed by those aged between 25–34 years (37.5%). Males (50.5%) were affected, more than females (49.5%) and doctors (26%) had more experiences of SSNIs than any other employees’ type, similar to findings by Makhoka [13] and Zafar et al. [43]. The “yes” responders also suffered more needlestick injuries (46.4%) than any other type of SSNI which is also consistent with previous studies [13, 45]. The devices most involved are syringe needles for intramuscular and subcutaneous injection (34.2%) and the major activity resulting in most needlestick injury was giving injections (30.9%) [11].

Across all employees, 43 (22.4%) suffered exposure through handling contaminated linen and waste. This exposure rate was high compared to that found in Gershon et al.’s [4] USA study in which only 10% of SSNI incidents were via contact exposure. Although this study and Gershon et al.’s [4] had different sample sizes, the higher percentage of employees exposed as a result of unsafe handling of waste/linen particularly, could be explained either as a sign of greater willingness to report incidents, or poorer compliance to appropriate PPE use.

Generally, the majority of SSNIs, 88 (45.8%) occurred four to five years ago; 58 (30.2%) happened two to three years ago with only 46 (24%) within the past year. This is indicative perhaps of a reducing incidence as a result of efforts by the hospital to improve safety within the workplace. This is in contrast to a national study which showed increasing cases of SSNIs [11]. The married (65.1%) and those with diploma qualification (26.6%) sustained more SSNIs perhaps due to solely to their greater presence within the workforce. Those that had worked >5 years (64.1%) within the organization had more SSNIs which may just be a consequence of the fact that they have been at risk longer than those that have been in healthcare for a shorter time. When comparing the frequency of SSNIs, 66.1% stated it had happened to them only once and 33.9% said more frequently than this. In 46.9% of SSNIs, Antiretroviral Viral (ARVs) for PEP had been administered. This compared to only 10.1% in the national study [11]. This is perhaps indicative of ease of access to PEP services in the study hospital compared to accessibility in other local healthcare providers.

The impact of SSNIs on QOL of HCWs was assessed using both the QOLS and symptoms scale scores for the “yes” responders (n = 192). The symptoms scores were processed and described in two levels as; those who were ‘most affected’ versus those ‘least affected’ to bring greater clarity to the data. The ‘most affected’ experienced the selected symptoms frequently, from more than once daily to more than once per week and their symptom scores were lower; ranging from 1 to 3 in the 6-point scale used. It revealed that just over half (52%) of HCWs felt depressed, suffered anxiety and worrying (51.6%) and were ‘despairing about the future’ (50%). The results are summarised as in Fig. 1. Gershon et al. [4] noted that 53% of their respondents had feelings of anxiety, 18% insomnia, 13% depression, 10% a loss of appetite, 10% sleepiness, and 10% frequently cried, confirming that it is psychological symptoms that most staff experience post-SSNI.

OPEN IN VIEWER

Fig.1

Percentage of HCWs “most affected “and “least affected” with symptoms burden following SSNIs (n = 192).

The QOLS scores were calculated as recommended by Burckhardt and Anderson for the affected 192 workers to show the impact on their QOL. The lower the scores were, the poorer the QOL and vice versa [33]. The total scores for HCWs ranged from 32 to 112 with a mean score of 79.2 (SD = 20.33). The mean score was less than the predicted 90 for a generally healthy population [33]. Noticeably, there were variations in QOLS means scores for HCWs who had different demographic variables (See Table 3). For instance, the mean scores for males were higher (81.4) compared to that of females (76.9). Similarly, the ‘married’ mean was 81.3 compared to the ‘singles’ score of 75.3, perhaps because the ‘married’ received better support (from partners) than the ‘singles’ did at the time of their SSNI incident. Notably though, it was HCWs who used ARVs who had the lowest mean score of 64.8 followed by those who had needlestick injuries, as opposed to any other type of SSNIs.

The expected pattern of results where lower symptoms scores corresponded with lower quality of life scores was noted. However, overall these scores were lower than the QOLS scores achieved in the Norwegian general population study [40] where analysis was conducted looking only at lower scoring respondents who had disease or health problems. This unwell group from Norway’s mean score was 79.6 (SD = 13.8).

To establish if there was any relationship between symptoms and QOLS score, a scatter plot was drawn with QOLS scores on the Y- axis as the response variable and symptoms scores on the X-axis (Fig. 3). The relationship between QOLS and symptoms scores was found to have a linear positive trend, r² = 0.567 which is a moderate and positive relationship. Furthermore, Pearson’s coefficient of correlation (r) was found to be equal to 0.753, which indicated a positive association between symptoms and QOLS scores. This confirmed that the level of symptoms experienced by HCWs was positively associated with their quality of life. The higher the symptoms scores were, indicating a lesser impact of symptoms, the higher the QOLS scores were, therefore, the better quality of life was for affected HCWs, and vice versa.

To further identify the level of QOL impact on each of the 16-QOLS items, the scores in the Likert scale ranging from “Terrible” as the lowest score (1) to highest “Delighted” (7) were calculated [33]. The initial results which are given in detail (See Fig. 2) were presented in three groups for better and clearer understanding. Those who scored from 1 to 3 in each item, were generally ‘dissatisfied’ indicating a poorer QOL. The second group comprised of those who had ‘mixed’ feelings and scored the midpoint on the Likert scale, which was 4. The last group was those who were ‘satisfied’. The ‘satisfied’ group scored from 5 to 7 and was assumed to have the better quality of life. In the 16-QOLS items, the greatest dissatisfaction expressed by HCWs (by 24.5% of respondents) was with the item ‘plans to have and rear children’ (Fig. 4). This was a unique finding for this study. Other researchers have employed this analysis yielding different results. For instance in a study involving Sickle Cell disease patients, work – related QOLS item had the lowest score because many participants were jobless because of their disability [39].

OPEN IN VIEWER

Fig.2

Percentage score of HCWs in each QOLS item following SSNI.

OPEN IN VIEWER

Fig.3

Relationship between symptoms and quality of life scores.

OPEN IN VIEWER

Fig.4

Percentage of HCWs responding “dissatisfied,” “mixed” or “satisfied” to the QOLS items following SSNIs.

In trying to establish if relationships existed between the independent variables of age, sex, professional cadre, experience, marital status, level of education and ARVs use and the QOLS scores of HCWs following SSNIs; results showed that SSNI type, ARV use, level of education and employee cadre generally had statistical association (p < 0.05) to QOLS scores. However, there were specific QOLS items which didn’t have any statistical significance. These are summarized in Table 4.

Finally, 43 HCWs responded to the open-ended question about their SSNI experiences. These were summarized into thematic areas and the majority of the comments, roughly half, were related to the impact of SSNIs on their lives. One HCW stated;

Fifteen responses were associated with service delivery relating to the management of SSNIs within the study hospital. Five were about safety standards and practice issues, such as the need to enforce compliance with good practice on waste management, the use of appropriate personal protective equipment (PPEs) and easy access to first aid facilities such as an eyewash sink in cases of splashes with body fluids. Three people responded regarding other more general issues. Nonetheless, for some, the SSNI incident was seen as a learning event. For instance, one staff member learned to think more about safety practices and commented,

The important role of the occupational health service was also recognized. For example, one respondent felt,

7.1 Limitations of the study

This study would have been aided hugely had a QOLS study been completed that examined the Kenyan general population. This study relied on the reference point mean score of 90 for a healthy population based on Anderson and Burckhardt’s initial view of the QOLS [32] but that may not be appropriate to a Kenyan context. According to Wahl et al. [40], comprehending and comparing any QOL indicators or measures in a study is helped if there is a wider general population reference comparison score. The ability to compare the QOLS scores of affected hospital staff with their peers outside the hospital would have helped to assist in showing just how much of an impact an SSNI was having on staff. Data collection by self-completion introduced a risk that respondents would either under-rate or over-rate their scores, although work by Bliven, Kaufman and Spertus [56] suggests that electronic collection of QOL data was just as reliable as any other method and was the preference for 45 (82%) out of the 55 people who participated this way in their study. The issue of the effect of other wider confounding variables that may influence QOLS scores was more problematic, particularly since the QOLS scores for HCWs who were not exposed to SSNIs were not collected.

It is important to note that the study hospital is a private facility with several safety and quality improvement initiatives in place, including employee health programs that can be accessed easily by HCWs who sustain SSNIs. This is atypical in Kenya, where the international standards adopted by this hospital may not be representative of other Kenyan health care facilities. As a result, further studies which involve local public healthcare facilities may have provided data that better reflects the wider national picture.