Table 2.
Description of study characteristics, reported outcomes, and narrative results for all seven studies retained from the systematic review
| Author | Intervention group | Instrument details | Sampling | Groups | Reported outcome | Results | Significant findings | Narrative description of results |
|---|---|---|---|---|---|---|---|---|
| Alhassan et al. (2016) | Supervision |
19 items 4-point Likert scale ranging from (1) "Very Disappointing" to (4) "Very Satisfactory" Reliability: Cronbach's alpha: >0.7 |
234 participants at endline (Intervention = 103, Control = 131) |
Group 1 (intervention): 32 facilities received SCE intervention Group 2 (control): 32 control facilities |
Motivation Factor 1: Physical work environment and resource availability |
Difference between Intervention and Control at Baseline: mean= −0.01 (SE= 0.01) Difference between Intervention and Control at Follow-up (2 years later): mean= 0.10 (SE=0.08) Difference in Difference: mean= 0.11 (SE=0.12) |
NS | Difference-in-difference analyses of the levels of staff motivation from 2012 to 2014 found that SCE-receiving facilities rated motivation proxies (career prospects, perceived workload, and overall motivation) higher than non-SCE facilities. (p<0.0001) The association between SCE and financial incentives and physical work environment was low or negative. |
| Motivation Factor 2: Financial and extrinsic incentives1 |
Difference between Intervention and Control at Baseline: mean= −0.19 (SE= 0.01) Difference between Intervention and Control at Follow-up (2 years later): mean= −0.09 (SE=0.08) Difference in Difference: mean = −0.28 (SE = 0.12) |
NS | ||||||
| Motivation Factor 3: Intrinsic incentivesb |
Difference between Intervention and Control at Baseline: mean= 0.22 (SE= 0.04) Difference between Intervention and Control at Follow-up (2 years later): mean= 0.44 (SE=0.07) Difference in Difference: mean = 0.22 (SE = 0.08) |
NS | ||||||
| Motivation Factor 4: Career prospects and opportunities for further education |
Difference between Intervention and Control at Baseline: mean= 0.68 (SE= 0.11) Difference between Intervention and Control at Follow-up (2 years later): mean= 0.79 (SE=0.10) Difference in Difference: mean=0.11 (SE=0.17) |
+ | ||||||
| Motivation Factor 5: Perceived workload and staff availability |
Difference between Intervention and Control at Baseline: mean= 0.51 (SE= 0.11) Difference between Intervention and Control at Follow-up (2 years later): mean= 0.51 (SE=0.08) Difference in Difference: mean=0.00 (SE=0.14) |
+ | ||||||
| Overall motivatione |
Difference between Intervention and Control at Baseline: mean= 0.24 (SE= 0.04) Difference between Intervention and Control at Follow-up (2 years later):mean= 0.32 (SE=0.06) Difference in Difference: mean=0.08 (SE=0.07) |
+ | ||||||
| Aninanya et al. 2016 | Compensation; Supervision |
14 items4-point Likert scale ranging from (1) "Strongly agree" to (4) "Strongly disagree" Reliability: Cronbach's Alpha: 0.7041 |
50 participants at endline (Intervention = 25; Control = 25) |
Group 1 (intervention): 6 facilities allocated as intervention-receiving PBI Group 2 (control): 6 facilities allocated as comparison – did not receive PBI |
Motivational construct: Intrinsic Motivation |
Difference between Intervention and Control at Baseline: Score= -0.25 Difference between Intervention and Control at Endline (3 years later):Score= 0.20 Difference-in-Difference:Score= 0.00 |
NS | Overall motivation was slightly higher in comparison compared to intervention group. Overall motivation increased in intervention group, but remained constant in comparison group over time. Difference-in-difference was small. All changes in motivation were not statistically significant. |
| Overall motivation5 |
Difference between Intervention and Control at Baseline: Score= -0.11 Difference between Intervention and Control at Endline (3 years later): Score= 0.08 Difference-in-Difference:Score= 0.01 |
NS | ||||||
| Carasso et al. 2012 | Compensation |
3 items 5-point Likert scale ranging from (1) "Very satisfied" to (5) "Very dissatisfied" Reliability: Cronbach's Alpha: 0.67 |
90 participants at endline (Intervention = 33; Control = 57) |
Group 1 (intervention): 14 facilities where user fees were removed- resulting in lower worker financing Group 2 (control): 6 facilities where user fees were still charged-resulting in higher worker financing |
Extrinsic motivation |
Mean motivation at endline for Intervention group: Mean= -0.55 Mean motivation at endline for Control group: Mean= 0.33 |
- | Staff in intervention group (with poorer health financing) reported higher extrinsic motivation than those in the control group. This finding was statistically significant (p<0.05). |
| Hosseinabadi et al. 2013 | Supervision |
17 items 5-point Likert scale Range not described Reliability: Stated as reliable; not described in paper |
40 participants at endline (Intervention = 24; Control = 16) |
Group 1 (intervention): 1 facility received quality circles Group 2 (control): 1 facility did not receive quality circles |
Motivational Factors |
Mean motivation at baseline for Intervention Group: mean= 35.25 (SD= 8.38) Mean motivation at endline (3 months) for Intervention Group: mean=43.50 (SD=7.63) Mean motivation at baseline for Control Group: mean=36.68 (SD= 8.51) Mean motivation at endline (3 months) for Control Group: mean= 34.25 (SD= 7.93) |
+ | There were no statistically significant differences in mean motivational scores between groups at baseline. At endline, motivation was significantly higher in intervention compared to control group. |
| Liu et al. 2017 | Supervision; Compensation; Lifelong learning |
10 items 4-point Likert scale ranging from (1)"Strongly agree" to (4) "Strongly disagree" Reliability: not described |
1050 participants at baseline 920 participants at endline |
Group 1 (baseline): 1 facility receiving life-long learning and compensation intervention Group 1 (endline): historical control – same facility did not receive life-long learning and compensation intervention |
Internal work Motivation |
Score at baseline:Mean= 3.15 (SD= 0.40) Score at endline: mean= 3.22 (SD= 0.64) Difference in score from baseline to follow-up (2-years later) not reported. Calculated by authors to be: mean= 0.07 |
+ |
Post intervention, there was a statistically significant positive change in internal work motivation over time (p<0.05). Internal motivation differed by the assigned work unit- with those from the operating room and post-anesthesia care unit demonstrating the lowest scores of internal work motivation. |
| Shen et al. 2017 | Compensation |
39-items 5-point Likert scale ranging from (1) "Very dissatisfied" to (5) "Very satisfied" Reliability: not described |
326 participants at baseline (Intervention = 147, Control 1 = 87, Control 2 = 92) 357 participants at endline (Intervention: 166, Control 1 = 92, Control 2 = 99) |
Group 1 (intervention): 10 districts received PBF Group 2 (control 1): 10 districts received EF Group 3: (control 2): 10 districts received no PBF and no EF |
Motivational Construct 1: Teamwork |
Pairwise regression results: PBF v. EF: β=0.39 SE=3.13 PBF v. Control: β=0.93 SE=1.43 EF v. Control: β=1.62 SE=3.51 |
NS | There was no statistically significant difference in any of the motivational outcomes between groups. |
| Motivational Construct 2: Autonomy |
Pairwise regression results: PBF v. EF: β=0.82 SE=4.31 PBF v. Control: β=1.31 SE=1.77 EF v. Control: β=1.30 SE=4.49 |
NS | ||||||
| Motivational Construct 3: Recognition |
Pairwise regression results: PBF v. EF: β=0.38 SE=3.28 PBF v. Control: β=-0.84 SE=1.33 EF v. Control: β=-0.89 SE=2.85 |
NS | ||||||
| Motivational Construct 4: Change |
Pairwise regression results: PBF v. EF: β=-2.10 SE=2.66 PBF v. Control: β=1.03 SE=1.24 EF v. Control: β=3.83 SE=2.64 |
NS | ||||||
| Motivational Construct 5: Self-concept |
Pairwise regression results: PBF v. EF: β=-0.73 SE=1.87 PBF v. Control: β=0.77 SE=1.08 EF v. Control: β=2.21 SE=2.36 |
NS | ||||||
| Motivational Construct 6: Work environment |
Pairwise regression results: PBF v. EF: β=-1.79 SE=2.60 PBF v. Control: β=1.26 SE=1.26 EF v. Control: β=4.31 SE=3.03 |
NS | ||||||
| Motivational Construct 7: Leadership |
Pairwise regression results: PBF v. EF: β=-3.08 SE=4.89 PBF v. Control: β=1.21 SE=2.61 EF v. Control: β=5.55 SE=5.15 |
NS | ||||||
| Motivational Construct 8: Well-being |
Pairwise regression results: PBF v. EF: β=1.10 SE=2.98 PBF v. Control: β=2.42a SE=1.24 EF v. Control: β=3.93 SE=2.50 |
NS | ||||||
| Vermandere et al. 2017 | Systems support |
23 items 5-point Likert scale ranging from (1) "Strongly agree" to (5) "Strongly disagree" Reliability: not described |
55 participants at baseline (Group 1 = 17, Group 2 = 16, Group 3 = 22) 40 participants at 1st follow-up (Group 1 = 12, Group 2 = 12, Group 3 = 16) 39 participants at 2nd follow-up (Group 1 = 10, Group 2 = 12, Group 3 = 17) |
Group 1 (intervention 1): 5 facilities receiving monthly evaluation Group 2 (intervention 2): 5 facilities receiving monthly evaluation + financial incentives Group 3 (control): 5 facilities receiving not receiving facility audits |
Overall motivatione |
Motivation at Baseline (Group 1): median=88.5 IQR=87–92 Motivation at Follow-up 2 (1 year later) (Group 1): median=90 IQR=88–90 Motivation at Baseline (Group 2): median=93 IQR=86.5–95 Motivation at Follow-up 2 (1 year later) (Group 2): median=87 IQR=83–90 Motivation at Baseline (Group 3): median=84.5 IQR=79–93 Motivation at Follow-up 2 (1 year later) (Group 3): median=87 IQR=83–90 None of the differences between 2nd follow-up and baseline were reported but p values presented were all non-significant |
NS | There was no statistically significant difference in motivation between intervention group and comparison group at baseline, endline, or over time |
Significant findings: + is positive,—is negative, NS non-significant findings at 95%
aFinancial and material work conditions of a job (e.g., salary increment, promotion, accommodation, etc.)
bInner joy and satisfaction derived from a job (e.g., societal recognition and respect; appreciation shown by clients, etc.)
cAs authors only provided a pooled sample size from baseline and follow-up surveys, reported sample size is the same at baseline and follow-up
dBaseline or control performance is the pre-intervention performance in the intervention arm, or if not measured, the end performance in the control arm
eComposite score of multiple constructs