This is an open-access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in JMIR mhealth and uhealth, is properly cited. The complete bibliographic information, a link to the original publication on http://mhealth.jmir.org/, as well as this copyright and license information must be included.
Nigeria contributes only 2% to the world’s population, accounts for 10% of the global maternal death burden. Health care at primary health centers, the lowest level of public health care, is far below optimal in quality and grossly inadequate in coverage. Private primary health facilities attempt to fill this gap but at additional costs to the client. More than 65% Nigerians still pay out of pocket for health services. Meanwhile, the use of mobile phones and related services has risen geometrically in recent years in Nigeria, and their adoption into health care is an enterprise worth exploring.
The purpose of this study was to document costs associated with a mobile technology–supported, community-based health insurance scheme.
This analytic cross-sectional survey used a hybrid of mixed methods stakeholder interviews coupled with prototype throw-away software development to gather data from 50 public primary health facilities and 50 private primary care centers in Abuja, Nigeria. Data gathered documents costs relevant for a reliable and sustainable mobile-supported health insurance system. Clients and health workers were interviewed using structured questionnaires on services provided and cost of those services. Trained interviewers conducted the structured interviews, and 1 client and 1 health worker were interviewed per health facility. Clinic expenditure was analyzed to include personnel, fixed equipment, medical consumables, and operation costs. Key informant interviews included a midmanagement staff of a health-management organization, an officer-level staff member of a mobile network operator, and a mobile money agent.
All the 200 respondents indicated willingness to use the proposed system. Differences in the cost of services between public and private facilities were analyzed at 95% confidence level (
This study demonstrates a case for the implementation of enrolment, encounter management, treatment verification, claims management and reimbursement using mobile technology for health insurance in Abuja, Nigeria. Available data show that the introduction of an electronic job aid improved efficiency. Although it is difficult to make a concrete statement on profitability of this venture but the interest of the health maintenance organizations and telecom experts in this endeavor provides a positive lead.
This study aims to document costs associated with and provide justification for adoption of mobile phone as an alternative to drive the uptake of community-based health insurance schemes (CBHIs). Studies have shown that one of the biggest challenges of health care systems in developing countries is financing [
On the basis of current health insurance coverage estimates, most clients settle payments for health services out of pocket at point-of-service [
The NHIS was setup to provide a regulatory framework to support the social health insurance system for Nigerians [
For the purpose of this study, all discussions referencing health insurance will refer to the CBHI hybrid focusing on maternal and child health. Various models of this scheme exist, one type, taking its operational model from its name, is organized and managed by community members through a committee sometimes called “Ward Development Committee”[
Maternal and child health domain of PHCs in Abuja, Nigeria, was the scope of this work. Treatment received at a referral center was excluded to reduce complexity.
Abuja is the federal capital territory (FCT) of Nigeria. Its population is 2.29 million [
There has been a massive growth in and adoption of mobile phone and related services in Nigeria over the last decade [
This research was the output of the principal investigator’s masters level dissertation studies. Data collection and analysis were conducted between December 2013 and February 2014. The Research Ethics Committee of the University of Liverpool, United Kingdom, approved this study.
This study is an analytic cross-sectional survey, which used a hybrid of mixed methods, involving key informant interviews supported by a “throw-away’ prototype software demonstration. A review of the current literature was conducted to guide the appropriateness of the approach. Health facility clients and health care providers were interviewed using structured questionnaires. The income and expenditure at both the private and public PHCs were analyzed. A client and a provider were interviewed in each health facility visited. Fifty public and 50 privately owned PHCs were considered representative of the study area. A throw-away prototype software system was designed to guide respondents’ understanding of the electronic enrolment and encounter systems. The prototype software conceptual framework is shown in
Prototype Software Conceptual Framework.
Two structured questionnaires were used for health facility clients’ (n=100) and the health workers’ (n=100) interviews. One hundred PHCs were randomly selected and visited by the research assistants. Fifty each of private and public primary health facilities were purposefully sampled to provide representative spread across public and private PHCs. The participant information sheet and the informed consent forms were read to the clients/health workers and health workers, and the consenting clients were interviewed after appending their signatures. One client and one health care provider were interviewed in each PHC visited. Interviewed clients had to be aged 18 years or older and must have received antenatal, immunization, or delivery service in the health facility visited. Client selection was on the first contact basis. Clients not meeting these criteria were excluded from the study, and the next available client was assessed and interviewed. The health worker interviewed was the most senior officer in the health facility at the time of visit by research interviewers. For confidentiality, all identifiable information was excluded from the questionnaire; to ensure data validity and reduce bias, research assistants and interviewers were enlightened on the objectives of the research and trained on how to ask and obtain answers for each of the questions to reduce bias and interobserver variation, as they were not supervised during the interviews. The health facility and LGA codes were adapted from the directory of the health facility [
The client questionnaire had 26 questions, which were grouped under 3 parts: A, B, and C. Part A of the client questionnaire contained questions that related to the reliability of the services in the health facilities. Information on the “time of last visit” was used to measure the frequency of health facility visits. Although the frequency of client falling ill and other factors might be beyond the scope of this research, these data were necessary to document costs necessary for good return on investment of any system targeting the primary health facility. Client’s perception of the service provided was captured using the “service rating” question. This was also validated using the “willingness to recommend others to the health facility,” question, as it is expected that a client would only be willing to recommend others to the service if it was above average by their rating. This might not necessarily be an accurate measure of quality, as other factors may have influenced the response.
Part B of the questionnaire captured the cost of services provided by asking questions about “amount spent” during the current visit, the service provided, and whether drugs were provided. Knowledge of NHIS and the willingness to enroll were also assessed in this section. Interviewers were trained to rephrase the NHIS questions if the need arose. Each interviewed client was asked about the premium they are willing to pay.
Part C only tested the current capacity of the client to use mobile technology and phone ownership. Clients where asked if they owned a mobile phone and how long they have owned and used a mobile phone. Their ability to send structured text messages using SMS was also assessed in this section of the questionnaire.
Similarly, the health worker’s questionnaire had 3 parts: A, B, and C. Part A asked questions to ascertain the reliability of service in the health facility. They were interviewed about their years of experience. They also responded to how satisfied they were with services provided in the PHC and how satisfied they perceive the community members are with the services provided. This section assessed the staff strength, which is a measure of the human resource capacity in a PHC. It has a direct bearing on the quality of service. Human resource is also a factor of the cost of providing service in a health facility.
Part B assessed the cost of services at the facility. The questions “drugs given,” “cost of antenatal care,” “cost of outpatient department,” and “cost of delivery” were all used to ascertain and document the eventual cost of services in the health facilities. Their awareness and knowledge of NHIS was also assessed in this section. Part C assessed the health worker’s phone ownership and capacity for using phone for SMS and other applications.
Similarly, key informant interview was conducted for a midmanagement HMO representative, an officer-level staff of a mobile network operator, and a mobile money agent.
The survey data were analyzed using Statistical Package for Social Sciences software (version 22) [
The responses were thematically analyzed in relation to reliability and sustainability indicators [
All 100-client respondents were expectant or new mothers; 51 (51%) were aged 18-34 years, 48 (48%) were aged 35-49 years, and only 1 (1%) was aged more than 50 years.
Only 25% of clients had visited a health facility in the last month, and 32 (32%) in the last quarter. Fifteen (15%) had visited a health facility within 6 months and 18 (18%) within the last year; 10 (10%) had not visited any health facility in over a year. Twenty-three (23%) of interviewed women had not been sick for over a year, and 26 (26%) were sick between 6 and 12 months.
Clients’ mean last clinic visit and last period of illness in months by age group.
Age group (years) | Mean last clinic visit (months) | Mean last period of illness (months) | n (%) |
18-34 | 4.12 | 6.48 | 51 (51) |
35-49 | 2.61 | 7.30 | 48 (48) |
≥50 | 24.00 | 1.50 | 1 (1) |
Total | 3.01 | 6.34 | 100 (100) |
The health facility staff strengths were also analyzed, as depicted in
PHC type and staff strength in Abuja (N=100).
PHCatype | 1-5 Staff |
6-10 Staff |
≥11 Staff |
N (%) |
Public PHC | 5 (10) | 7 (14) | 38 (76) | 50 (100) |
Private PHC | 6 (12) | 20 (40) | 24 (48) | 50 (100) |
Total | 11 (11) | 27 (27) | 62 (62) | 100 (100) |
aPHC: primary health center.
Service rating by clients.
The monthly expenditures incurred by PHCs were grouped into personnel, operations, fixed equipment, and medical consumables costs. The personnel expenditure primarily included clinical and nonclinical staff time; fixed equipment costs; and covering equipment such as couches, beds, building, and so forth. The operation cost was further subdivided into power, water, paper printing, and transportation costs. The private PHCs are entirely self-funded for all categories of expenditure, whereas their public equivalents have personnel cost completely covered and receive unstructured government subsidy for operations. The level of subsidy depended on various factors such as the proximity of the health facility to the client, client load, and even political interests. In other cases, some of the public health facilities do not receive subsidies.
On the other hand, the health facility income stream was analyzed based on responses from clients and health workers. The results of the analyses are summarized in
Service costs as reported by health facility staff interviewed.
PHC ownership | Cost of antenatal service (₦) |
Cost of OPDaconsultation (₦) |
Cost of delivery service (₦) |
|
Public | ||||
n | 50 | 50 | 50 | |
Mean | 444.5 | 435.5 | 426.5 | |
Standard deviation | 541.2231 | 539.7704 | 538.1601 | |
Private | ||||
n | 50 | 50 | 50 | |
Mean | 2691.44 | 2008.46 | 2691.44 | |
Standard deviation | 1472.0359 | 1538.7633 | 1472.0359 | |
Total | ||||
N | 100 | 100 | 100 | |
Mean | 1567.97 | 1221.98 | 1558.97 | |
Standard deviation | 1578.7391 | 1393.1766 | 1584.7048 |
aOPD: outpatient department, PHC: primary health center.
The cost of service as reported by the interviewed health workers, shown in
Amount spent for health service on clinic visit day.
PHC ownership | Amount spent at clinic today (₦) |
Premium willing to pay (₦) |
|
Public | |||
n | 50 | 50 | |
Mean | 1691.42 | 2380.270 | |
Standard deviation | 1953.861 | 1736.8265 | |
Private | |||
n | 50 | 50 | |
Mean | 4291.04 | 2380.350 | |
Standard deviation | 2663.122 | 1427.1472 | |
Total | |||
N | 100 | 100 | |
Mean | 2991.23 | 2380.310 | |
Standard deviation | 2665.777 | 1581.4980 |
Similarly, client responses to amount spent for health service as shown in
As shown in the figure, the clients interviewed at the public PHCs spent less than those at private PHCs. On the interview day, 21 clients (41%) in public versus 1 (2%) in private PHCs reported spending between ₦ 100 and ₦ 500. The other extreme also shows that half of the clients (n=25; 50%) at private clinics indicated spending more than 5000 on the interview day, whereas only 4 clients (8%) reported doing the same in public clinics.
The client interview data for premium affordability and willingness to pay for families were marginally different from health provider interviews, as shown in
Services provided by PHCatype.
Service provided | Private PHC (%) | Public PHC (%) | n (%) |
Antenatal attendance | 25 (50) | 23 (46) | 48 (48) |
Immunization attendance | 12 (24) | 24 (48) | 36 (36) |
Delivery attendance | 13 (26) | 3 (6) | 16 (16) |
aPHC: primary health center.
Estimates provided in
Software setup and operation expenditure.
Year | Seed fund expense | Estimate ($) |
Year 1 (Immediate) | One-time software setup | 29,800.00 |
One-time infrastructure cost and cloud computing setup cost | 50,000.00 | |
Yearly support and maintenance | 10,000.00 | |
Training software administrator | 31.25 | |
Other personnel costs | 20,000.00 | |
Total estimated seed fund for software | 109,831.25 | |
Year 1 | Annual PHCaexpense | |
Annual data and SMS ($18.72 × 2 × 12) | 450.00 | |
Mobile device (10″ tablet) | 562.50 | |
Spare mobile device | 562.50 | |
Annual device replacement (20% device value) | 112.50 | |
Annual electricity for device charging | 300.00 | |
Training of 5 PHC staff members | 156.25 | |
Total year 1 cost per PHC | 2,143.75 | |
Years 2, 3, 4, and 5b | Annual data and SMS ($18.72 × 2 × 12) | 450.00 |
Annual device replacement (20% device value) | 112.50 | |
Annual electricity for device charging | 300.00 | |
Total 1-year recurrent cost per PHC | 862.50 | |
4 Year Recurrent cost | 3,450.00 | |
Total 5-year cost per PHC | 115,425.00 |
aPHC: primary health center, SMS: short messaging service.
bYearly recurrent cost.
Excluding drugs and other medical supplies, other annual expenditure expected per PHC for personnel cost and operation costs detailed in
Personnel and operation expenditure of a primary health centera.
Serial number | Monthly expenditure | Unit annual rate ($) | Number of personnel | Amount ($) |
1 | Clinical and nonclinical personnel | 3,750 | 10 | 37,500 |
2 | Operation costs | 1875 | ||
Total | 39,375 |
aExcluding medical consumables in a digitized primary health center, as it has no effect in this cost-documentation exercise.
Operational costs summarized in
The key informant interview with a midmanagement HMO representative indicates that they reimburse for services using a combination of fee-for-service and capitation within 30 days of claims submission. The 30 days allows for vetting and verification. The most common reason for delayed reimbursement was “missing or incomplete detail.” The main reason for client service denial was attributed to misunderstanding of service level availability for selected plans and list of hospitals including those for referrals. They noted that enrollees complain most about perceived “substandard drugs” given as against out-of-pocket payees. Health facilities are currently reimbursed through bank wire irrespective of location. This HMO indicated that the most basic package for a family of 4 per annum costs ₦90,000 ($452). When enquired if they encourage monthly premium payments, the response was no. On the other hand, an individual will have to enroll with a premium in the range of ₦20,000-450,000 ($100.5-$2261.3), depending on service coverage and hospital selection.
Although a formal interview with a representative of a mobile network operator could not be conducted or because of several reasons, in an informal interview, an officer level staff explained that it is possible to enroll for health insurance using SMS or unstructured supplementary service data and that it is a part of business priority interest for his organization. Similarly, a mobile money agent interviewed to assess the viability of premium payment and service reimbursement to health facilities through mobile money indicated that mobile money business was neither lucrative nor widespread in reach to support the enterprise.
Comparison of private and public service expense cost by health facility.
Premium that the interviewed clients are willing to pay per family per annum.
The major findings from this study are that there is wide disparity in cost of payment for health services between private and public PHC institutions in Abuja, Nigeria. The results show that the deployment of a mobile-supported scalable insurance scheme will require significant investment to set up and operate. The interviews indicate that although services are not at an optimal level, clients are generally happy with the service they currently receive. This, we believe, may be related to their experience and exposure. The results of a recent survey conducted by Sambo et al in Kaduna State in north-western Nigeria were consistent with the results of this study on maternal neonatal and child health service costs [
The key informant interview with the HMO showed that the current premium pricing may not be realistic, as only 10% (n=10) of clients interviewed were willing to pay a cumulative of ₦20,000 ($100.5) per family per annum. The HMO reportedly charges this minimum per individual and about ₦100,000 ($502.5) per family, when we consider that many public health facilities may still need a significant initial investment to meet certain service delivery quality standards that will drive demand for health services. The need for alternate and seed funding becomes important for improving both infrastructure and quality level and to subsidize the premium pricing.
In February 2013, a leading Nigerian newspaper, the
The Nigerian government has provided a minimum of 1% of its consolidated national funds for health care through the National Health Law 2014 [
During this study and analysis, the cost of health services at the referral facility (secondary and tertiary health facilities) was assumed to be the same as in the primary health facility and the mean computed from the interviews. However, this is not always so, as the treatment increasingly becomes expensive as we move from primary care to secondary to tertiary. The income analyses were conducted using out-of-pocket payment extrapolation. The actual proposed income profile will vary slightly based on insurance enrolment and buy-in by each community. According to Abuja FCT, millennium development goals office, drug supplies to these public health facilities have often been inadequate [
Mobile health insurance presents an opportunity for wider expansion of insurance adopters. It was easy to establish that a mobile enrolment system will improve efficiency and reliable. The interests demonstrated by mobile network operators and health management organizations demonstrate business interest and willingness to participate. However, profitable cost for identified stakeholders would hover around the private PHCs’ mean service costs, still requiring significant subsidy. This means that to achieve universal health coverage, insurance costing model must consider identified variations. This study successfully demonstrated that mobile supported enrolment, encounter management, treatment verification, and claims management and reimbursements can be efficient and sustainable. It also shows that using technology can aid in accountability and thus reliability of CBHI financing and reimbursement. This study successfully documented income and expenditure associated with personnel, fixed equipment, and operation, as they influence the adoption of mobile insurance-management system. Costs for medical consumables have been a topic of many other researches and were not considered in this study.
community-based health insurance
federal capital territory
Global System for Mobile Communications
health maintenance organization
local government area
National Health Insurance Scheme
primary health center
None declared.