Attachment II QnA Analyzing Feedback Loops DI Report.pdf

Prepared under the Democracy and Governance Analytical Services Indefinite Quantity Contract, #DFD-I-00-04-00229-00

Submitted to:

USAID/Serbia

Prepared by:

MARCH 2016

This publication was produced for review by the United States Agency for International Development. It was prepared by Democracy International, Inc.

ANALYZING FEEDBACK LOOPS THAT

STRENGTHEN DEMOCRATIC

ACCOUNTABILITY

WITHIN PARTNER COUNTRY

SYSTEMS

FINAL REPORT, LIBERIA

DISCLAIMER

This is an external report. The views expressed in this document are the authors' and do not necessarily reflect the views of the United States Agency for International Development or the United States Government

Submitted to:

USAID/Washington

Prepared by:

Chris Demers Gerald Hyman

Contractor:

Democracy International, Inc.

7600 Wisconsin Avenue, Suite 1010

Bethesda, MD 20814 Tel: 301-961-1660 www.democracyinternational.com

ANALYZING FEEDBACK LOOPS

THAT STRENGTHEN DEMOCRATIC

ACCOUNTABILITY

WITHIN PARTNER COUNTRY

SYSTEMS

FINAL REPORT, LIBERIA

MARCH 2016

http://www.democracyinternational.com/

TABLE OF CONTENTS

LIST OF ACRONYMS I

FINAL REPORT 1

Introduction: Genesis and Focus of Research 1

Major Findings 2

Health Structures and Process Supporting Feedback 9

Ministry of Internal Affairs Structures and Processes Supporting Feedback19

Incentives, Motivations, and Issues of Actors Involved In Feedback 20

Factors Affecting Feedback Success 26

Observations on Decentralization and Governance 36

ANNEX A: LIST OF OBSERVED COMMUNITY HEALTH CONCERNS A-1

ANNEX B: LOFA COUNTY MEETING SCHEDULE B-1

ANNEX C: COMMUNITY-LEVEL CASE MANAGEMENT REPORTING FORM C-1

ANNEX D: PBF INDICATORS D-1

ANALYZING FEEDBACK LOOPS THAT STRENGTHEN DEMOCRATIC ACCOUNTABILITY – FINAL REPORT 1

LIST OF ACRONYMS

BPHS Basic Package of Health Services, the fundamental collection of health care services to be supplied by a health facility

CBO Community Based Organization, a group of community members (usually) loosely and informally organized to undertake collective actions to assist one another and improve the quality of life. Normally less structured than an NGO, without full-time staff, and less focused on a particular activity or problem.

CDC The County Development Council, with representatives from all of the districts, meets once a year under the chairmanship of the County Superintendent to allocate (or recommend the allocation of) the county’s (donor-provided) development budget which augments the line ministry-supplied budgets (including salaries for example).

CDSC County Development Steering Committee, composed of county representatives of the line ministries, including County Health Officer, under the chairmanship of the County Superintendent, organized to create and implement policies and programs for the development of the county as a whole.

CHC Community Health Committee consists of five to nine members elected by the community for a given catchment area. Each CHC chooses one of its members to serve on a Community Health Development Committee (CHDC)

CHDC Community Health Development Committee consists of members of CHCs elected to represent at the clinic level, chaired by the Officer in Charge of the clinic, intended to provide community input and help hold the health staff accountable to the public

CHO County Health Officer: the senior health official in a county, responsible for its health, especially the functioning of its health facilities and its health staff, and reporting to the Ministry of Health in Monrovia.

CHB County Health Board consists of the County Health Officer (representing the County Health Team) as well as other members of the county administration, representatives of civil society, and representatives of health partners. Chaired by the County Superintendent, it provides input from other parts of the county administration and county citizens, not just the health staff.

CHSC County Health Steering Committee, consists of the county health board and representatives from the county and district administrations, county health teams, district health officers and partners (NGOs, civil society organizations, or technical partners, oversees the implementation of the PBF scheme at the county level.

CHT County Health Team (headed by the County Health Officer and including (were it fully staffed) consisting of nurse(s), nurse midwife, certified midwife, finance

ANALYZING FEEDBACK LOOPS THAT STRENGTHEN DEMOCRATIC ACCOUNTABILITY – FINAL REPORT 2 officer, procurement officer, data clerk, surveillance officer, reproductive health officer, social worker, environment officer, mental health officer, logistics officer, carpenter, plumber, driver, etc.

CHV Community Health Volunteer, a local, unpaid citizen assisting in the delivery of health services in a community

CHW County Health Worker: the MoH paid health staff at the county level.

CM Certified Midwife: professionally trained, degreed midwives in contrast to the Traditionally Trained Midwives

CS County Superintendent the senior administrative official of a county, reporting directly to the Ministry of Internal Affairs in Monrovia and, inter alia, chair of the County Health Board

CSO Civil Society Organization, non-governmental groups comprised of citizens normally dedicated to particular public interests or purposes.

DC District Commissioner the senior administrative official of a district, reporting the County Superintendent and, inter alia, chair of the District Health Board

DHB District Health Board consists of the District Health Officer (representing the District Health Team) as well as other members of the district administration, representatives of civil society, and representatives of health partners. Chaired by the District Commissioner, it is intended to provide input from other parts of the district administration and district citizens, not just the health staff.

DHIS District Health Information System, the health data representing the health of a district reported to the Ministry of Health and used to measure progression and regression of health care as well as a national picture of the health of Liberia’s citizens

DHO District Health Officer: the senior health official in a district, responsible for the district’s health, the functioning of its health facilities and its health staff, and reporting to the County Health Officer

DHT District Health Team, headed by the District Health Officer and (were it fully staffed) consisting of nurse(s), nurse midwife, certified midwife, finance officer, procurement officer, data clerk, surveillance officer, reproductive health officer, social worker, environment officer, mental health officer, logistics officer, housekeeper, laundry workers, carpenter, plumber, driver, etc.

EPHS Essential Package of Health Services, an augmentation of Basic Package of Health Services to be supplied by health facilities

FARA Fixed Amount Reimbursement Agreement, agreement between the Government of Liberia and the Government of the United States by which the USG will compensate (“reimburse”) the Government of Liberia for providing certain health care services in Bong, Lofa, and Nimba counties and containing both the

ANALYZING FEEDBACK LOOPS THAT STRENGTHEN DEMOCRATIC ACCOUNTABILITY – FINAL REPORT 3 performance measures to be met by the GoL and the corresponding compensation by the USG

GAC General Auditing Commission (GAC), the independent supreme audit institution of Liberia headed by its Auditor-General.

IAA Internal Audit Agency, a unit of the government of Liberia charged with auditing GoL entities.

gCHV General Community Health Volunteer, a volunteer sanctioned by the MoH who lives and works directly in a community and receives training to assist with reporting and management of community health issues and care.

G2G Government-to-Government: the form of donor assistance where funds are given directly to partner government, in this case Liberian government. In the case of health, the FARA is the mechanism for G2G.

GOL Government of Liberia

Health Facility health clinics, health centers, and hospitals providing health care to a “catchment area” consisting of a population inhabiting a specific region or zone

HMIS Health Management Information System, the national collection of detailed health information developed and implemented by the Ministry of Health and based on the regular input of health data from communities through districts and then counties.

INGO International NGO. International organizations that support governments and donors in development.

IP or Partner Implementing Partner, the INGOs or local CSOs that support the GoL in the delivery of health care

MoH Ministry of Health of the Government of Liberia

MIA Ministry of Internal Affairs of the Government of Liberia

NDS National Drug Service: organ of GoL that manages the availability of drugs to the Health Facilities, including the procurement, monitoring, and delivery of drugs from the NDS warehouses to the various facilities upon receipt of requisitions from health facilities within their allowable drug budgets.

NGO Non-Governmental Organization: an organization of citizens in the social space between families and the state.

NHP National Health Policy describes goals, rules and guidelines governing the tiered system of health delivery, including the services and management expected at different levels of care

OIC Officer in Charge, the senior professional Ministry of Health staff member heading a health facility

ANALYZING FEEDBACK LOOPS THAT STRENGTHEN DEMOCRATIC ACCOUNTABILITY – FINAL REPORT 4

PBC Performance Based Contracting: a contract whose terms and (usually compensation) depends on the execution of agreed-upon benchmarks with the normal proviso that exceeding those benchmarks will result in increased compensation.

PBF Performance Based Financing, a mechanism under the FARA under which the Ministry of Health provides service goals and metrics for measuring them and under which bonuses are provided to facilities which exceed their service/performance goals.

TTM Trained Traditional Midwives local volunteers, not government employees and not compensated by the government, who have been trained by their predecessors to help deliver babies in their local communities.

WHO World Health Organization, a specialized agency of the United Nations concerned with public health

ANALYZING FEEDBACK LOOPS THAT STRENGTHEN DEMOCRATIC ACCOUNTABILITY – FINAL REPORT 1

FINAL REPORT

INTRODUCTION: GENESIS AND FOCUS OF RESEARCH

USAID commissioned research in 2013 looking at feedback opportunities for government services in Liberia and Ghana, where USAID is supporting government agencies in delivering those services through a government-to-government (G2G) assistance agreement. The goal of the research is a greater understanding of how USAID can promote domestic accountability when it designs G2G programming. In Liberia, the G2G agreement covers health care in three northern counties—Bong, Lofa, and Nimba—so the research meant examining the feedback loops affecting the Ministry of Health in delivering its basic package of health services in those counties.

In early 2014 a scoping trip was conducted to map feedback opportunities in the health sector, attempting to look for “bright spots” to center the research around. A second, longer and more detailed trip was to complete the research a few months later. However, those three counties (plus Monrovia itself) constituted the epicenter of the Ebola pandemic originating on their borders with Guinea and Sierra Leone and the research was postponed until the epidemic ended.

In 2016, once the epidemic was contained, a joint USAID and Democracy International research team conducted the final deep-dive which examined closely activities surrounding some of those bright spots, including community and facility-level health structures and traditional actors. The researchers looked not only at what these structures and actors are intended to do, but how the system actually works and the motivations that incentivize action and the challenges that inhibit their effectiveness.

Over a period of three weeks, the research team visited offices of the Ministry of Health, Ministry of Internal Affairs, National Drug Service, Internal Audit Agency and several health partners in Monrovia. Field research took place in Lofa and Bong counties where the team interviewed County Health Teams, Superintendent Offices, District Health Teams, District Commissioners, traditional leaders, health facilities, health volunteers, civil society organizations, and citizen groups. Most of the meetings with citizens and facility staff were conducted in focus groups although some were conducted with individuals as well. The meetings with district and county officials and with the staff of Africare and International Rescue Committee (the contracted managers of health care) were typically with individuals as were the meetings with ministry staff in Monrovia. In total, well over [100] people were interviewed in this fashion. In addition, the team reviewed relevant official documents.

The report provides substantial detail governing the ideal and actual provision of health care as well as the incentive structure for its delivery. Some details do not relate directly to the feedback loops but offer context—that helps explain strengths and weaknesses of feedback loops.

Liberia is a very poor country. Bong, Lofa, and Nimba are among its poorest and most vulnerable counties. Whatever the deficiencies, which are well known throughout the health care system, including to Ministry officials in Monrovia, the effort to provide health care is often heroic. Health care providers are struggling with very rudimentary resources to meet the needs of their patients.

ANALYZING FEEDBACK LOOPS THAT STRENGTHEN DEMOCRATIC ACCOUNTABILITY – FINAL REPORT 2

They recognize fully the gaps between what they would like to provide and what they are able to provide. Similarly, the standards which officials would like to attain, especially those in Monrovia, are advanced for Sub-Saharan Africa.

No doubt, there are many citizen complaints about the provision of health care. A common concern is the distance which some have to travel to their nearest health facilities: in the rural areas of Bong, Lofa, and Nimba. Not infrequently, the lines at the facilities are so long that they cannot be examined before closing and are required to return the following day. Often they have no way to do that beyond asking total strangers to take them in. Often the health facilities do not have the medications they need and are presumed to have. Sometimes staff do not show up, although that complaint was less often heard in the 2016 deep dive than in the 2014 scoping trip, in part because Ebola has made urgent health care staffing.

MAJOR FINDINGS

1) Ministry of Health (MoH) officials at the highest level, including the Minister of Heath, are sufficiently aware of the strengths and weaknesses of the entire health system.

Officials understand patient concerns and needs at every level of the system—the local facilities, the health centers, the health clinics, the hospitals, the district and county health teams, and the Ministry of Health in Monrovia—so there is little if any new and incisive information that could flow but is not flowing to the top decision-makers. The only exception is the incidence of outbreaks of communicable diseases, of which Ebola was the most well-known, but also including measles, dysentery, cholera, and the like. The value of information flowing up through the system to the Ministry lies in the capture of data illuminating the health situation in the country on a regular basis and showing improvement or deterioration. New and better information could provide additional nuance perhaps but is unlikely to change, let alone correct, the fundamental understanding of senior officials. In fact, between the many reports (weekly, monthly, and quarterly) that flow upward to the county health teams and to Monrovia and the personal inspection visits central officials pay to the counties, districts and facilities, the Ministry of Health is aware of all of the problems and issues detailed below. Their inability to address the many problems of the health system stems from lack of resources, incentive structures, governance inadequacies, and actionable information, not the absence of information altogether.

2) However, the reverse is quite the opposite. The flow of information from higher to lower levels of the health system—whether ministry to county, county to district, district to health facility, and (perhaps most important) facility to patient—is weak. Indeed one of the regular complaints at each level is the lack of information each official or patient receives from the level immediately higher. The exception at each level consists of health policies, campaigns, and information dispensed from the Ministry of Health about the treatment of outbreaks of communicable diseases, Ebola being again the most dramatic and recent example. In a country as poor as Liberia and with such relatively inaccessible places, the knowledge about public health problems is impressive. Put differently, the

ANALYZING FEEDBACK LOOPS THAT STRENGTHEN DEMOCRATIC ACCOUNTABILITY – FINAL REPORT 3 feedback loop is weak, not because the information flow up from communities through the system is weak—it is remarkably robust for such a poor country—but because the “feedback” down is poor.

3) The line of authority on health issues through the Ministry of Health structure is more robust than the one through the Ministry of Internal Affairs officials. Health officials are part of two government networks: one consisting of Ministry of Health officials; the other consisting of Ministry of Internal Affair’s officials. In theory, the two are integrated in the county and the district. For example, in those counties that have fully developed district health teams, the district health officer reports up to the country health officer but also, along with other “line” or service officials (like education) and the district’s own administrative officials, to the district commissioner who is part of the Ministry of Internal Affairs. The same is true one level higher at the county level: the county health officer reports to the Ministry of Health in Monrovia but also, along with all the district commissioners and various county-level officials, to the county superintendent. The district commissioner is theoretically responsible for the administration of every aspect of his district, including all of the line ministries’ functions like health; the county superintendent is similarly responsible for the administration of the entire county. In practice, however, the line of authority through the service or line ministries is stronger than the one through the Ministry of Internal Affairs officials to the line ministry officials.

The county health officer reports primarily to the Ministry of Health and only secondarily to the county superintendent. The same holds at the district level. In part, the vertical reporting strength and accountability is a remnant of the former vertical structure of administration, integrated only at the presidential level. It is also a reflection of the weaker interest, technical expertise, and responsibility of the county superintendent and the district commissioner in health care, compared to their interests in and responsibility for the general administration of, respectively, their county or district. For that same reason, the authority of the district commissioner or county superintendent over health care and their accountability for health care is weaker than those of the respective health officers. “Horizontal feedback” is weaker than “vertical feedback”.

4) De-concentration in the health system has primarily reached the county level, but no lower. The Government of Liberia is decentralizing, or more technically “de-concentrating”, authority from Monrovia to lower levels of government on the principle of “subsidiarity”: transfer authority to the most local level that can handle it effectively.

In the area of health, the de-concentration has reached the county level, but not much lower. Budgets and therefore decisions about prioritization of alternative resource dispositions have now been de-concentrated from Monrovia to the county level but not yet to the district level, let alone below to the local facilities. The officers in charge (OICs) of the local health facilities and the district health officers to whom they report are asked to propose annual work-plans but without costing them out and without the ability to reprioritize them when, inevitably, the resources available do not cover all of the proposed work or requested needs. That prioritization is done at the county level and subsidiary units are instructed about which of their work-plan elements have been approved and to what extent. Based on work-plans submitted by the local health facilities

ANALYZING FEEDBACK LOOPS THAT STRENGTHEN DEMOCRATIC ACCOUNTABILITY – FINAL REPORT 4 and aggregated and modified by the district health team (DHT), the county health team (CHT) drafts and submits to Monrovia a fully budgeted work-plan. The funded (actually, underfunded) work-plan and budget are returned by the central ministry in Monrovia to the county health team for re-prioritization based on the actual, approved work-plan and budget. The districts and, in turn, the facilities are informed about which parts of their work-plans have been funded, but just as they did not create a budget in passing the work-plans up to the county health team, they do not re-prioritize the elements in their approved work-plans. They simply implement the CHT decisions. So de-concentration and subsidiarity have yet to find their way to the district, facility and community level.

5) Stock-outs and a private drug store system have reduced the health system’s ability to gather complete information regarding illnesses contracted and drugs used by Liberian citizens. A private drug store “system” coexists with the government health system, which theoretically provides free health care to every Liberian. For example, although patients are entitled to free medicines in many cases, many patients patronize local medicine shops (or pharmacies or apothecaries) or purchase medicines from private purveyors. Why would Liberians pay for what they are entitled to receive free of charge, especially when almost all Liberians are poor? Because, in many cases, the free, public system promises more than it delivers. Too often medicines theoretically free of charge are simply not available at the government facilities: there are “stock-outs” for a variety of reasons. In fact in many cases, patients at free clinics are given prescriptions and told to fill them at the pharmacies. Among other consequences, the MoH is unclear about the full range and scope of illnesses, how they are treated, how the government’s treatment regime fits with that of the private purveyors, and what the national health care system looks like as a whole. A perception remains that if you pay money you get better treatment in health system. Though government facilities are free-of-charge, this perception seems to include both private and public facilities. Apart from their effects on the health of the purchasers of privately sold medications (The purveyors will sell whatever their customers demand and can pay for irrespective of the effects—or non-effects—of the medications on the health of the customers.), the private market distorts the conclusions the MoH can draw about Liberian health drawn since these private sales are not reported to the Ministry. That distorts the information feedback loop. (Of course, every system with a private sector carries that same problem for public health statistics.)

Some of these stock-outs are due to inadequate monitoring and auditing of the warehouses and local storage containers. Some are due to corruption, or what the Liberians call “siphonage”, in which government-purchased medications are drained off to private dealers and sold to the public. To say the least, not all of the pharmaceuticals ordered and paid for by the government are actually delivered and certainly not to the local facilities.

6) Patients experience insufficient feedback about diagnostic test results. The outcome of insufficient feedback about the results of these diagnostic tests leaves the patient uncertain of the diagnosis, the prognosis, and the course of treatment required to alleviate their illness. It leaves the patient with little agency over his or her health care.

Moreover, it deprives the health care workers of the ability to learn from their

ANALYZING FEEDBACK LOOPS THAT STRENGTHEN DEMOCRATIC ACCOUNTABILITY – FINAL REPORT 5 misdiagnoses and improve their diagnostic skills and care. In effect, the results of the diagnostic tests play a much more limited role in the treatment and prevention of illness than they could. A more complete feedback loop with all of this information available to each critical player in the treatment loop could improve results with potentially small additional effort.

7) The unpaid volunteers—the traditionally trained midwives (TTMs) and the community health volunteers (CHVs)—bear much of the brunt of ensuring health care in the more inaccessible area. Unpaid volunteers play a critical role in getting patients to a health care facility and reporting on the health profile up from community to government and instructions and policies down from government to citizens. They are motivated to do so primarily out of a concern for their communities and neighbors. CHVs do get some training, the odd attendance at workshops, and the potential for some advantage when a salaried government position becomes available. The TTMs get the worst of all possible deals: they are unpaid; they are asked to escort pregnant women long distances through difficult terrain just days before giving birth; they are implored by the pregnant women to provide food or places to stay once they reach the health facilities; they are theoretically entitled to about USD $15 for doing so, but those legally responsible for the payments, the husbands or boyfriends of the women, often do not pay and even deny paternity; moreover, the TTMs (and the pregnant woman) are fined if, except in emergencies, the pregnant woman gives birth in the community rather than at the facility.

These volunteers perform their duties at great risk and with no return. There are multiple disincentives to serve the community and few strong incentives other than personal commitment and humanitarianism to do so. Some TMMs described their personal motivations with the words, “we cannot watch the new babies or their mothers die”. This volunteer arrangement may diminish, which is why the median age of the TTMs we met is almost certainly above 50 years. These are primarily old women with decreasing energy and resources. In addition to the economic benefits of their free labor, these volunteers are the eyes and ears of the government health system at the most basic level. They know who is sick and with what ailments. They carry much quantitative and qualitative public health knowledge. They are the sources of information about disease and death. For example, they keep the ledgers on which the Ministry of Health relies for its data about the condition of Liberia’s population.

8) In practice though not in theory or law, the relation between the health officials (i.e., GoL health staff) and the implementing INGOs remains too murky and authority is still unclear and diluted. The Government of Liberia (GoL) in general and its county governments in Bong, Lofa, and Nimba counties, now run their health care system with donor funding. In theory, the INGOs that previously managed the health system under contracts to and with grants from donors now provide the essential or basic package of health services (EPHS or BPHS) under performance-based contracts for the Government of Liberia in general and its county governments in particular. In short, through a Fixed Amount Reimbursement Agreement (FARA) mechanism in these counties, USAID funding goes directly to the government (G2G) and through it to the INGOs. However, in practice, the county governments do not fully supervise the INGOs and the local facilities as well

ANALYZING FEEDBACK LOOPS THAT STRENGTHEN DEMOCRATIC ACCOUNTABILITY – FINAL REPORT 6 as the population still look to the INGOs for direction and resources. To perhaps a lesser extent, so too do district health officials and even the county officials.

9) The performance based financing (PBF) program bonus incentive structure remains unclear to community-level actors but has promise. The government has instituted a national performance-based financing (PBF) program to incentivize better health performance, particularly by the local facilities. In most counties, the PBF funds come from a multi-donor pool of general development funding which is often inadequate for actual PBF bonuses. However, in Bong, Lofa, and Nimba, FARA funds are actually available from USAID through the MoH to provide bonuses to facilities that meet or exceed their performance targets. In brief, 20% of the bonus goes to the implementers (the County Health Team and the INGOs), 65% of the remaining bonus goes to augment staff salaries and 35% of the remaining bonus is available for use by the community to provide health-related benefits1. The facility-related community decides on how that 35% is used.

Commonly, it is used to upgrade maternal health resources like places for pregnant women to stay; staff dormitories; fencing; and even sometimes micro-economic development programs. Facility portion of bonuses often go toward bumping up staff salaries. While the entire PBF system, including the targets and the mechanisms of decision, is clear in Monrovia, it is more vague at lower levels. Community members are often unclear about the feedback relation between performance and bonus. They think of any bonus, like they do of the health care system more generally, as a gift from the government but without clarity about why it was given, or what period of time they are being rewarded for. Consequently, the intended incentive feedback is weak. Incentivizing performance through bonuses depends on an understanding of the behavior the bonus is designed to encourage, so the objectives of PBF would be enhanced by a broader, clearer, more public understanding of the performance targets that will be rewarded. No doubt the indicators of performance are many (see Annex D) and should perhaps be prioritized to incentivize more acutely the most important performance (although the points allocated differ for the various indicators and so should incentivize the most important performance). But prioritized or not, they will not create concrete incentives if they are unknown to all actors. It would also be enhanced by a clearer and more public procedure for deciding on the use of the 35% community share of any bonus.

10) Among the most frequent complaints of the health workers is low pay or no pay at all, and the mechanism of payment. Many health workers have difficulty meeting their basic expenses, including those with government salaries. But in addition to criticisms of the rate of compensation, are those directed at the method of its delivery. Typically, pay is automatically deposited at the local bank, but in counties like Bong and Lofa, banks are located only in Gbarnga and Voinjama respectively. Liberia remains predominantly a cash economy. Each person is required to travel to one of these towns to collect his or her salary. That is a trip of several hours each way, usually on a local bus. Moreover, in the

1 “The remaining 35% of the earned bonus can then be used for innovative activities that are likely to improve the performance towards achieving a set goal.” Ministry of Health and Social Welfare; Performance Based Financing Operational Manual; March 2012; p. 42.

ANALYZING FEEDBACK LOOPS THAT STRENGTHEN DEMOCRATIC ACCOUNTABILITY – FINAL REPORT 7 not infrequent event that the deposit has not arrived or there is some problem with its deposit, the staff member is required to make the same trip two or three times. The financial cost of these trips is borne by the staff member of course, but the health cost is shared by patients. The staff member’s time away from the facilities decreases the facility’s ability to care for patients, contributing to the public’s annoyance at having to travel hours to get to a facility whose staff is absent. Almost every paid staff member has a phone. Given the long distances, the poor roads, the scarcity of public transportation, and the commonality of electronic transfers and payments for private transactions like cellphone payments, it seems almost inconceivable that the payroll system still relies on personal trips to distant banks, especially given the frequency of deposit errors and multiple trips. A mobile banking system would decrease that annoyance and increase the availability of staff to care for patients. Incidentally, it would almost surely improve the Ministry’s own payment system and record-keeping.

11) Among the more minor staff criticisms of the health care system is the large number of meetings and other bureaucratic requirements required as part of their responsibilities.

Hardly unique to Liberia or its health care system, still the number of groups and meetings seems excessive, especially for the facilities which are hard-pressed to meet the needs of their constituents. As noted elsewhere, patients complain about long waiting lines and the frequent necessity to return the following day to be examined, in part because the health care workers are in meetings, sometimes with the district health teams and away from the facilities themselves. Health centers and clinics have even more meetings to attend but they also have larger staffs. Eliminating meetings with relatively little impact on health delivery would leave time for more attention to feedback.

12) Although not directly involved in the delivery of health care, traditional leaders like the hierarchy of chiefs (town chiefs, clan chiefs, paramount chiefs, and supreme traditional chiefs) play a role, especially in the feedback loops. These leaders are particularly important in the delivery and explanation of healthcare messages to ordinary patients, the mobilization of citizens for health care like vaccinations, and the response to health emergencies like Ebola and other communicable diseases. They have immediate connection with local health volunteers like the traditionally-trained midwives. For example, residents of rural areas can be reluctant to share their personal health problems “with strangers” like the health facility, center, clinic, and hospital staff but are much more willing to do so with the chiefs who live among them, have grown up with them, and are woven into their social networks. These are kinsmen and neighbors, people of trust, not strangers. During the Ebola crisis, in particular, these were the people who persuaded villagers to modify traditional burial practices, one of the vectors of the disease, and to accept the recommendations of and treatment by the healthcare staffs.

13) The MoH recognizes that qualitative information about health services is not captured adequately by government or the PBF program. The researchers were told there are plans for a Qualitative Management Unit in the MoH, the demand for this partly driven by the Ebola response. The PBF program is planning to introduce qualitative indicators along with quantitative indicators in its reporting framework, which will help measure facilities more holistically in their performance. In its current monitoring and evaluation

ANALYZING FEEDBACK LOOPS THAT STRENGTHEN DEMOCRATIC ACCOUNTABILITY – FINAL REPORT 8

(M&E), the PBF program does incorporate “supportive” supervision visits that are intended as a quality assurance measure. Though there has been limited implementation of a client satisfaction survey under the PBF program, this is expected to be a bigger piece of PBF 2.0. This addition should create a vehicle for measuring health service quality as perceived by the customer. One facility the researchers came across was using a suggestion box to help capture qualitative information (among other feedback that could be received). The anonymous box was opened at the CHDC meeting with both community members and facility staff present. It is unclear how much the suggestion box is used given high illiteracy rates.

14) While there are certainly some weaknesses in health-system feedback loops, there are also a number of “bright spots” in which feedback works efficaciously. For example, the collection of data by volunteers at the local level and staff in the facilities seems to be quite diligent (although no attempt was made to follow either staff or volunteers or to test their reported data against some more rigorous recording method). Moreover, there was little difference among participants in the elucidation of concerns and problems in health care. For example, there was no real discrepancy between the complaints or concerns expressed by local volunteers and those expressed by the staff of the health clinics or health centers. Indeed the most senior officials in the Ministry of Health were quite aware of the concerns and complaints expressed by the patients and volunteers. The information and data flow from the local level through the entire system seems strong and has an obvious impact on officials as it makes its way up through the reporting system. Third, the value placed on information and explanation is strong throughout the system as manifested, for example, in the insistence by Ministry of Health officials on implementing and improving the Health Management Information System (HMIS) and the importance also attached to the thorough collection of accurate information by local staff and volunteers. Reciprocally, at least certain messages from the top are heard clearly as they move down through the system. Two examples are illustrative. Most women now give birth in a health facility, the result in part of the information campaign initiated by Monrovia about high infant and maternal mortality rates resulting from giving birth at home under the supervision of traditionally-trained midwives. Of course the fines which are now levied against TTMs who do not bring their patients to the clinics and even against mothers who give birth in their own homes clearly incentivizes birthing in health facilities. But unless pregnant women were aware of the fines and their reasons, they would not have abandoned home-delivery. So the information is clearly flowing down as well as up. Similarly, the response to Ebola which required, for example, the abandonment of strong customs surrounding funerals and touching others illustrates the efficacy of the Ministry’s information campaign about Ebola and the measures needed to contain it.

HEALTH STRUCTURES AND PROCESS SUPPORTING FEEDBACK

Communities have a number of avenues to provide feedback on health services. Many of these avenues are practiced simultaneously as community members are uncertain what will work, who

ANALYZING FEEDBACK LOOPS THAT STRENGTHEN DEMOCRATIC ACCOUNTABILITY – FINAL REPORT 9 needs to know, and who might ultimately be responsive. In rural areas, General Community Health Volunteers (gCHVs) are key health interlocutors with the public, listening to community grievances, collecting epidemiological information, referring clients to facilities, and sometimes performing case management. They are appointed by government based upon recommendations from the community, often through Community Health Committee (CHC) members, and are an extension of government health services. The Community Health Services Policy, and to some extent the Basic Package for Health and Social Welfare Policy (BPHS), explain their duties. GCHVs collect both technical and non-technical information from the public. CHCs and Community Health Development Committees (CHDCs) are also present in rural areas and are defined by the Community Health Services Policy. They are civil society structures encouraged by government to monitor and hold accountable health services. Their members are elected by the community. CHC members are lay people who generally have not accumulated technical skills in health. CHDCs are simply CHC members from different catchment areas who represent the CHC at the facility level. Communities also feed information through CHCs though the information passed is generally less technical. TTMs are found across rural areas and urban areas as well.

TTMs, like gCHVs, are skilled health volunteers, though their focus is entirely on maternal and child health. They are often the recipients of feedback on health services from pregnant and nursing mothers, and sometimes women more broadly.

The following graphic from the 2011 National Community Health Services Strategy explains an ideal feedback arrangement once a community based information management system is in place:

Figure 1. Information system flow from community to center and the feedback mechanism Each of the three categories of health volunteers mentioned has a monthly meeting with the nearest health facility. Each group has their own meeting with the facility and sometimes they also have joint meetings. These meetings are a focus for channeling community health feedback into government, which means clinics, the lowest level facility in rural areas, are often at the fore for collecting health feedback from the public. Community members do frequent health clinics, health centers and hospitals and thus they have the opportunity to give feedback directly to facilities. Community members prefer to give feedback directly to facilities whenever possible, but given that many do not live close to facilities, concerns might only be recognized when away from the facility. Moreover, given the limited attention facility staff have for individual patients, it is not always the case that community members can give desired feedback directly to facilities.

In addition, since health care is free, it is seen as a kind of “gift” from the government. Poor patients are diffident in most societies when confronting government officials, and when they believe they are receiving a gift, they are even less likely to criticize or complain. Finally, they are wary of irritating the health staff on whom they rely for current and future care, about whom they remain somewhat suspicious, and who are “strangers” to them, i.e. not born and raised in their communities. Still, they do complain and the health staff is quite aware of the complaints so in that respect the local feedback loop is fairly strong. In urban areas where hospitals exist and facilities are much closer to patients, community members can give higher-level facilities direct feedback. However, sometimes community outreach efforts of hospitals are not as strong, ANALYZING FEEDBACK LOOPS THAT STRENGTHEN DEMOCRATIC ACCOUNTABILITY – FINAL REPORT 10 especially when they are private hospitals. Concerns and questions to facilities are wide ranging and can be on technical and non-technical matters.

Community members sometimes give health feedback to the traditional system of chiefs directly.

This feedback is usually non-technical and done when community members have grievances or feel they were treated unfairly. In urban areas where gCHVs do not exist, chiefs are more frequently an intermediary with the health system. Chiefs generally acknowledge their limited knowledge of health services. On occasion, technical matters are brought to the attention of chiefs when community members are uncertain where else to turn. Chiefs generally do not respond to such technical concerns, but instead find the right people for community members to speak with.

Sometimes feedback from community members is brought directly to partners supporting the government in implementing health services. Partners do not openly solicit feedback from community members, but they are often seen as the ones with resources, which is why some community members bring concerns to them. The researchers found that local civil society organizations (CSOs) are not frequently subject to feedback by community members. They are usually not involved in health service delivery and frequently critique government health services in an arms-length relationship, Thus it is not surprising that community members do not see them as a helpful place to deliver feedback on health services. Some local CSOs were previously involved in client satisfaction surveys of health services. If the GOL decides to implement such surveys again, CSOs may be viewed by community members as a location to deliver feedback.

As compared to the most distant county officials, members of District Health Teams (DHTs) are sometimes the recipients of community member feedback, as is the District Commissioner from time to time. DHT’s offices are usually located at a facility, which is part of the reason they may be recipients of community feedback. They are also present in communities when doing surveillance or case verification which avails them to the public.

In terms of visibility, health volunteers cannot see beyond county-level actors. They are unaware of what gets passed further than the county level. Facilities, even clinics, do have some visibility

ANALYZING FEEDBACK LOOPS THAT STRENGTHEN DEMOCRATIC ACCOUNTABILITY – FINAL REPORT 11 beyond the county level as they deal directly with national-level laboratories and the National Drug Service (NDS) directly.

Although many of the feedback channels discussed are exhibit obstacles to effectively sending responses back down the channel, especially those that need higher level involvement, there are places where rapid feedback occurs. No process highlights this better than the “swabbing” process set up as part of the Ebola response which is completed upon all deaths in Liberia. Each deceased person gets swabbed and the specimen is rapidly taken to one of a few laboratories. If the result is positive, a team is rapidly deployed to the community and strict protocols are put in place to trace the disease and isolate other people if potentially affected. If persons are not found positive for Ebola, the community and facility producing the swab hear nothing. While the absence of response may be readily understood as a negative test result, it is a significant source of anxiety for affected communities and facilities. With something as serious as Ebola, it would seem even negative cases should receive a response. This reality speaks to the absence of feedback that is common in Liberia. People do not think to tell people about things that affect them. Facilities not only complained about Ebola, they complained about all testing and referrals that are carried out by higher facilities, where response is often absent, leaving them uncertain of a condition or outcome, and less able to diagnose and treat patients in future.

In addition to swabbing for Ebola death, the health system also has a standing practice of DHTs or Community Health Teams (CHTs) auditing all maternal deaths. This practice, though not as systematized and possibly not as regular as swabbing, shows that the health system can offer consistent feedback on causes of death.

Figure 2. Responsible actors for information flow in health services

ANALYZING FEEDBACK LOOPS THAT STRENGTHEN DEMOCRATIC ACCOUNTABILITY – FINAL REPORT 12

According to observations of the researchers, these are standard government actors and pathways responsible for health information sharing.

FEEDBACK CHANNELS UP

GCHVs pass a number of types of information into the health system. There are three different ways that gCHVs can report an incident to a facility: through a form, through a phone call, or passing a message with someone going to the facility. gCHVs also keep ledgers, which include information about community health conditions. The researchers were given a copy of a checklist that apparently some gCHVs use in their community data collection. The impression given to researchers was that this community health information gets reported to the facility through the focal person, but does not make it into the Health Management Information System (HMIS) or any reports that go further. Plans apparently exist for a separate Community Health MIS where this information will be input and centrally collected. In most communities, gCHVs travel to the nearest facility weekly to meet with the focal person and join at least one monthly meeting at the facility.

In rural Liberia, clinics are the locus for feedback because of their proximity to communities and the frequency with which they are used. Clinics handle a variety of information that they either attempt to address or feed up to other health system actors. There are two overall types of information being passed up by health works: information about health concerns in the population and information about challenges in the local health system. Clinics pass information they receive from community members and health volunteers about disease outbreaks, challenges facing health campaigns, concerns about an individual’s treatment at health facilities, ANALYZING FEEDBACK LOOPS THAT STRENGTHEN DEMOCRATIC ACCOUNTABILITY – FINAL REPORT 13 and referral requests to higher level facilities along to DHTs. The researchers were told that on occasion referrals are “not honored” by higher-level facilities. The volume of referrals should be kept in mind, apparently only a small percentage of cases are too complicated for clinics to address. Clinics pass to the DHT epidemiological and other statistical information through the HMIS, which in turn goes to the CHT and MoH.