The Cost of Intramedullary Nailing for Femoral Shaft Fractures in Dar es Salaam, Tanzania

Abstract

Background

Femoral shaft fractures are one of the most common injuries seen by surgeons in low- and middle-income countries (LMICs). Surgical repair in LMICs is often dismissed as not being cost-effective or unsafe, though little evidence exists to support this notion. Therefore, the goal of this study is to determine the cost of intramedullary nailing of femoral shaft fractures in Tanzania.

Methods

We used micro-costing methods to estimate the fixed and variable costs of intramedullary nailing of femoral shaft fractures. Variable costs assessed included medical personnel costs, ward personnel costs, implants, medications, and single-use supplies. Fixed costs included costs for surgical instruments and administrative and ancillary staff.

Results

46 adult femoral shaft fracture patients admitted to Muhimbili Orthopaedic Institute between June and September 2014 were enrolled and treated with intramedullary fixation. The total cost per patient was $530.87 (SD $129.99). The mean variable cost per patient was $419.87 (SD $129.99), the largest portion coming from ward personnel $144.47 (SD $123.30), followed by implant $134.10 (SD $15.00) medical personnel $106.86 (SD $28.18), and medications/supplies $30.05 (SD $12.28). The mean fixed cost per patient was $111.00, consisting of support staff, $103.50, and surgical instruments, $7.50.

Conclusions

Our study provides empirical information on the variable and fixed costs of intramedullary nailing of femoral shaft fractures in LMICs. Importantly, the lack of surgical capacity was the primary driver of the largest cost for this procedure, preoperative ward personnel time. Our results provide the cost data for a formal cost-effectiveness analysis on this intervention.

Appendix: equations and assumptions

See Table 5.

Table 5 Processes comprising time and motion analysis

Disposable supply cost details

The median supplier cost given in the database was used in this analysis with the exception of midazolam for which only a buyer median price was reported. For medications packaged in large multiuse vials, the total volume per unit was divided by the volume used and multiplied by the unit cost to assess the cost used in each procedure. The unit cost of gauze, scalpels, and needles were obtained from MOI. The cost of single-use X-ray materials were estimated from a 2008 study in Botswana [32]. The unit cost of a 500 mL transfusion was taken from a 1999 study in Tanzania [33]. The cost per litre (CPL) of oxygen gas produced from a central oxygen generator using grid electricity was estimated using the parameters of a 2010 cost study in Papua New Guinea and Eqs. 3 and 4 [34]. The cost of halothane or isoflurane used in each procedure was calculated with Eq. 5. For each patient, the costs of each supply used was summed to give a total supply and medication cost per patient.

Equation 1: Surgical personnel costs per patient (CPP)

Salaries were separated into nine occupational categories (attending orthopaedic surgeon, anaesthesiologist, medical resident, nurse anaesthetist, nurse, radiology technician, or hospital assistant):

$${\text{Total}}\;{\text{surgical}}\; {\text{CPP}}\; = \; \mathop \sum \nolimits ({\text{Person}}\; {\text{hours}}\; {\text{category}}_{i} \times {\text{wage}}^{a} )$$

(1)

where a is the mean wage of ward staff, calculated by dividing the mean annual salary by 1840, the product of 8 h days and 230 working days per year (365–104 weekend days–15 sick days and 16 holidays).

Equation 2: Ward personnel CPP calculated for general and private wards separately

$${\text{Ward}}\;{\text{personnel}}\;{\text{CPP}} = \frac{{\# {\text{Workers}}_{1} + \# {\text{workers}}_{2} }}{2} \times \frac{{{\text{Wage}}^{a} \times 24 \;{\text{h}} \times {\text{LOS}}^{b} }}{{\# {\text{Patients}}^{c} }},$$

(2)

where $\# {\text{workers}}_{1}$ is the number of staff during normal hours (6AM–6PM), $\# {\text{workers}}_{2}$ is the number of staff during off-hours (6PM–6AM), a is the mean wage of ward staff, b is the patient length of stay, and c is the number of patients in ward.

Equation 3: Oxygen capital cost per litre (CPL)

$${\text{Capital}}\;{\text{CPL}} = \frac{{\left( {{\text{Oxygen }}\;{\text{generator}}\;{\text{cost}} + {\text{building cost}} + {\text{piping costs}}} \right) + \left( {{\text{Maintenance}} \times 10\;{\text{years}}} \right)}}{{{\text{Total}}\; {\text{litres}}\; {\text{produced }} 10\;{\text{years}}}}.$$

(3)

Equation 4: Oxygen electrical cost per litre (CPL)

$${\text{Electrical CPL}} = \frac{{{\text{Running}}\;{\text{power}}^{a} \times 0.63^{b} \times \$ 0.07/{\text{kWh}}^{c} }}{{{\text{Maximum }}\;{\text{hourly}}\;{\text{capacity}}\;({\text{L}})^{d} }}{\text{}}$$

(4)

where a is the (([{#Beds  × 0.75} + {all ORs × 10}] × peak hours/week × 60 min × 52 weeks) + ([{#beds × 0.75} + {emergency OR × 10}] × non-peak hours/week × 60 min × 52 weeks)) × 10, b is the capacity factor based on litres needed during peak/non-peak hours, c is the TANESCO price/kWh for high power users (December 2014 prices), and d is the assuming system is designed for operation at calculated peak need +10 %, (6000 L/h).

Equation 5: Liquid volume of halothane and isoflurane used in anaesthesia

$$V_{l} = \frac{{Q_{{{\text{O}}_{2} }} tMP}}{22.4\rho },$$

(5)

where $Q_{{{\text{O}}_{2} }}$ is the oxygen flow rate (L/min), P is the percent by gaseous volume of the compound, M is the molecular weight of the compound (g/mol), and t is the time (min) and $\rho$ is the density of the compound (g/mL) at 20 °C.

Equation 6: Femur fixation-specific instrument set costs per patient (CPP)

$${\text{Instrument}}\;{\text{CPP}} = \frac{{{\text{Instrument }}\;{\text{set }}\;{\text{cost}} \times \# {\text{sets}}}}{{{\text{Procedures}} \;{\text{per}}\;{\text{year}}^{a} \times {\text{lifespan}}}},$$

(6)

where a is the Femoral shaft fixation procedures per year.

Equation 7: Administrative and ancillary staff costs per patient, patient-day equivalent (PDE) method

$$\begin{aligned} \left( {\text{i}} \right) {\text{Total}}\;{\text{PDEs}} = {\text{Inpatient}}\;{\text{days}} + \frac{{{\text{Outpatient}}\;{\text{visits}}}}{{{\text{Inpatient}}\;{\text{equivalence}}\;{\text{ratio}}^{a} }} \hfill \\ \left( {\text{ii}} \right){\text{Femur}}\;{\text{PDEs}} = \frac{{\# {\text{Patients }} \times {\text{LOS}}}}{\text{Total PDE}} \hfill \\ \left( {\text{iii}} \right){\text{Cost}}\;{\text{per}}\;{\text{patient}} = \frac{{{\text{Total}}\;{\text{staff}}\;{\text{cost }} \times {\text{Femur PDEs}}}}{{\# {\text{Femur}}\;{\text{cases}} }} \hfill \\ \end{aligned},$$

(7)

where a is the inpatient equivalence ratio (IER) built by averaging the IERs of three previously published cost studies (3, 3.77, 4) [35–37].