“I prefer a male nurse to a female nurse”: patients’ preference for, and satisfaction with nursing care provided by male nurses at the Komfo Anokye teaching hospital

The underlying theory adapted for this study is the middle-range theory by Swanson [45]. In this theory, Swanson described the prevailing factors or actions which helps to foster positive patient outcomes [45, 46], restores [47] and unifies both patient and nurse [48] relationship. He defines care as “a nurturing way of relating to a value other toward whom one feels a personal sense of commitment and responsibility” [45]. This makes caring an integral component of nursing [49] due to the need for nurses to be self-knowledgeable and engrossed in ameliorating their patients’ discomfort [50, 51]. The middle-range theory is made up of five levels of care which helps to guide the actions of nurses, in which Level 1 examines the characteristics of caring persons; level II: commitment to providing caring; level III: the conditions that enhances or diminishes the likelihood of care being implemented by the interaction of the following variables (nurse, patient and organization); Level IV summarizes caring actions and level V are the consequences of caring [52]. This can be summarised as follows: knowing, being with, doing for, enabling and maintaining belief. The phenomena Kristen Swanson addressed reflects the process of nursing practice which could be best applied to this study where caring was the pivot upon which male and female nursing care processes were mirrored (viewed) in relation to patient’s satisfaction of care provided by male or female nurses. In holistic nursing care, the nurse must establish a rapport with the patient to enable her to get the patient’s cooperation. The nurse’s relationship with the patient provides basis for identifying the patient’s problems. The nurse through the caring process, plan with the patient and family, set short- and long-term goals to address the patient’s problem identified. Despite the use of these levels in previous nursing care studies, little attention was given to the incorporation of Level III in these studies [13]. This study builds on level III of the middle-range theory as used by Adeyemi-Adelanwa and colleagues [13] by assessing the factors which enhances or diminishes the likelihood of care provided by male nurses. Respondents demographic characteristics were used as predictors or independent factors for assessing their preference and satisfaction with male nursing care. It is important to indicate that preference and satisfaction are mutually inclusive and as such, using the same predictors has positive implication on the study findings. The theory’s effectiveness in strengthening nursing practice by improving patient satisfaction and outcome [46], and factors which are likely to diminish nursing care by male nurses in the future, made it useful for this study. Considering the negative label attributed to males in the nursing profession [17, 18] and the cultural orientation of society towards male nurses [21, 22], it is envisaged by this study that patients’ demographic characteristics could potentially facilitate or inhibit their preference for, and satisfaction with nursing care provided by male nurses.

The study was carried out at both the medical and surgical wards of the second largest hospital in Ghana, Komfo Anokye Teaching Hospital (KATH), Kumasi. KATH doubles as a major referral hospital serving mostly the regions in the northern part of Ghana [53]. The specific wards where the research was carried out were C1 (female ward), C2 (male ward), D3 (male ward) and D5 (female ward) of the surgical and medical wards respectively. C1 and C2 are on the first and second floor of the C-block respectively, whereas D3 and D5 are on the second and fourth floor of the D-block respectively. The blocks are linked by corridors. Each ward has a bed capacity of thirty-six (36). An inferential cross sectional study design, in which data were collected at a single point in time was used. The design preference was necessitated by the need to identify the prevalence of a condition, disease or a phenomenon among an identified population [54‐56]. In this case, prevalence looks at the percentage of the population who prefer and are satisfied with nursing care provided by male nurses.

A context based data on the opinion, preference and satisfaction of patients with nursing care provided by male nurses was used in this study. Data were collected within two weeks after the ethical approval on 20th March 2017. The underlying objective of the questionnaire [57] was to provide a breadth of understanding on patients’ estimation of the activities of male nurses within Ghana’s Health Sector, as well as provide a baseline information for prospective male nurses in Ghana. The need to identify respondents who meet certain practical criteria, are within accessible geographical proximity, and are available and willing to participate in the study, led to the use of convenience sampling [58, 59]. The sampling technique has been proven effective in health-related studies [60]. The technique is guided by the principle of homogeneity and as such, its results are not different from a study which made use of random sampling techniques [61]: it ensures that knowledge gained is representative of the population [62].

Going by this technique, 150 patients from the medical and surgical wards of KATH who were eighteen (18) years and above and had been admitted at least twenty-four (24) hours or more prior to the study were sampled for inclusion [63]. The draw back to this sampling technique is the presence of outliers and biases in the sampling procedure which could potentially undermine the representativeness of the study findings [64, 65]. This drawback was addressed by means of approaching all respondents who had been admitted in the aforementioned wards and only those who agreed to participate in the study were included. This approach was used as a criteria for checking biases and outliers since controlling all factors which could result in biases or outliers was neither possible in practice nor wise [66].

Data were analysed using Predictive Analytics SoftWare (PASW) version 16. The study instrument established an alpha value of 0.834. Differences in frequencies among respondents’ demographic characteristics by gender were assessed using χ² test analysis. Participants’ opinion about male nursing care were analysed using Mann Whitney U test due to the skewed distribution, and ordinal nature of the opinion parameters [67, 68]. We used ordinal logistic regression (OLR) to assess respondents’ satisfaction with male nursing care owing to the ordinal nature of the dependent variable. Respondent’s demographic characteristics was regressed upon their satisfaction with male nursing care, to calculate the crude odds ratio (OR). The good-of-test of the OLR was judged using Pearson’s, − 2 Log Likelihood Ratio and Test of parallel lines values. Respondent’s preference for male nurse was ascertained using logistic regression. Respondent’s basic information were all regressed upon participants’ preference for male nurse to calculate the crude OR. The statistical parameters; Hosmer-Lemeshow test and Nagelkerke R Square, were used to judge the model fit of the analysis. All statistical result were deemed significant at p < .05.

Ethical clearance was sought from both the medical and surgical directorates of the Komfo Anokye Teaching Hospital and the ethical review committee of the School of Medical Sciences, KNUST on 20th March 2017, with identification number CHRPE/AP/181/17. Besides this, oral consent were sought from the study participants after presenting an introductory letter from the Department of Nursing, KNUST, which detailed the essence of the study. Only participants who agreed to participate in the study were included. Confidentiality and anonymity of response were strictly adhered to; no names were recorded on the questionnaires and there was no alteration of the information obtained.

Respondents’ biodata by sex are presented in Table 1. Majority (108, 72%) of the respondents were below 45 years, married (58, 39%), self-employed (78, 52%) and had attained basic education (80, 53%). Overall, 86% of the study participants professed Christian beliefs and 72% were from the Akan ethnic group. Besides age, religious beliefs and ethnic groups, there was statistically significant differences between the male and female respondents based on the Pearson Chi-Squared test conducted. This proves the representativeness of the study sample (p < 0.05).

Table 2 presents a Mann-Whitney output on patients’ perception about male nurses based on their gender. The mean ranks for each group indicates the perceptual difference between the male and female patients. Primarily, more females had been attended to by male nurses than their male counterparts during the hospital visits. The female patients had higher mean ranks than males in the following variables: male nurses as polite and courteous, skilful with their craft and provided a comforting atmosphere when taking care of their patients. However, the male patients scored higher mean ranks than the females regarding the professional conduct and cooperation with of male nurses. Despite these differences in mean score, just two opinion variables: politeness and courtesy and comfortability with male nurses, were statistically significant between the male and female patients.

From Table 3, two of participants’ demographic information; marital status and religious affiliation, were significant predictors of their preference for a male nurse. Particularly, those who were single had an 88.9% chance [OR = 0.111, 95% CI (0.013–0.928)] of preferring a male nurse on any other visit to the hospital as compared to the married patients. Also, respondents who professed Islamic faiths were about 3553% more likely to prefer a male nurse [OR = 36.533, 95% CI (2.116–630.597)] on any other visit as compared to the Christians. The predictive values explain roughly 44.3% of the variation in the dependent variable and the Hosmer-Lemeshow result proves fitness of the model (p > 0.05).

From Table 4, respondents’ religion and educational level were significant predictors of their satisfaction with nursing care provided by male nurses. Specifically, respondents who were affiliated to a religious sect (OR = 2.347, 95% CI [0.076–1.630]) and educated (OR = 1.387, 95% CI [0.040–0.615]), had higher odds of falling in one of the higher categories of satisfaction with nursing care provided by male nurses as against the lower categories.