Improving the quality of child health services: participatory action by providers

Abstract

Objective. To test a quality improvement approach called COPE^® (Client-Oriented, Provider-Efficient services), for use in strengthening health systems and supporting Integrated Management of Child Health (IMCI) efforts.

Design. Pre- and post-intervention observations of client/provider interactions, facility audits, staff and client surveys, and focus groups to evaluate differences between eight COPE intervention and eight matched non-intervention facilities after a 15‐month intervention in 2001.

Setting. Primary care clinics in Guinea and Kenya.

Study participants. Health care providers and child caregivers.

Interventions. Over 15 months, the intervention supported four COPE exercises at each intervention site, supported supervisor training in quality management, and organized minimal training in topics selected by site staff as areas where training was needed.

Main outcome measures. Differences in staff’s and child caregiver’s knowledge, attitudes, and practices; differences in the quality of services provided.

Results. On almost every quality indicator (over 65 indicators), whether reported by staff, observed by evaluators, or reported by clients, the intervention sites performed statistically significantly better than control sites. Intervention sites were cleaner and more pleasant, with more respect and information for clients, and more privacy. Staff had better personal communication skills, better diagnostic skills, and prescribing practices and gave better home care instructions to carers. Clients in intervention sites were more informed and more satisfied, and their children had better immunization coverage than those in control sites.

Conclusion. COPE is a simple process, yet our study confirms that it can have a very dramatic effect on the quality of services. This study demonstrated how all areas of quality can be addressed by empowering health care providers to take action by using COPE. We suggest that COPE can complement Integrated Management of Childhood Illness (IMCI) training and can help to achieve better health for children.

child health, health services, participation, quality improvement

In recent years, new strategies for improving the diagnosis and holistic treatment of childhood illness, including the Integrated Management of Childhood Illness (IMCI), which focuses on the five major childhood diseases [1], have been introduced in many developing countries. Early results of IMCI performance show that clinicians have had considerable success in correctly identifying certain conditions [2–4]. However, many studies have also noted that there is insufficient emphasis on improvement of health systems or on how quality improvement can be operationalized [5].

COPE^®1, which stands for Client-Oriented, Provider-Efficient services, is a participatory, problem solving approach that invites health facility staff to identify and prioritize quality of care problem areas and encourages them to find their own solutions and subsequent actions to improve health services. Simple to implement and low cost, COPE was pioneered by EngenderHealth in the early 1990s and has been adapted by a variety of agencies working in low-resource settings at hospital and health center levels.

Because past experience had demonstrated anecdotally that COPE could be easily adapted to address a range of reproductive health service areas, several agencies with child health mandates (UNICEF and USAID) asked EngenderHealth to adapt COPE for use in child health services. To determine whether such a process could and should be scaled up, these agencies provided primary technical inputs and/or financial support to EngenderHealth’s Research and Evaluation Unit to conduct this study, the first systematic research to assess the effectiveness of COPE in improving service quality. This article presents the findings of a prospective, quasi-experimental study examining the effects of COPE^® for Child Health [6] on improving the quality of health services at the facility level.

The intervention

The COPE process is based on four simple tools and the belief that staff from all areas of a health facility should participate in problem identification and resolution, from administrative and support staff to service providers from wards and clinics. The main COPE tool is a set of self-administered guides with trigger questions organized within a framework of client rights (to information; access; informed choice; safety; privacy and confidentiality; dignity, comfort, and expression of opinion; and continuity of care) and health care staff needs (for facilitative supervision and management; information, training, and development; and supplies, equipment, and infrastructure)—all elements defined in this study as key components of quality of care at the facility level [6]. The questions help facility staff to systematically and consciously look at issues related to service quality, some of which may have never considered before (such as waiting times or client perceptions). The client exit interview tool encourages staff to talk with and listen to their clients about the quality of the services offered. The client-flow analysis tool measures how long clients wait for services and how much contact time they have with service providers. The action planning tool helps staff to identify root causes of their self-identified problems and to develop a realistic, time-bound plans that assign responsibilities to individuals. The whole exercise occurs several hours each day over a 2- to 3-day period when the clinic is quietest and when clients are not too inconvenienced. Action plans are reviewed at the next COPE exercise, three to four months in the future, when the COPE reflection–analysis–action cycle begins again. In the interim period, staff work to implement their action plans.

The COPE study in Guinea and Kenya occurred over a 15‐month period and was comprised of four COPE exercises interspersed with the implementation of staff action plans. In the first month of the project, external facilitators oriented district supervisors and intervention site managers to COPE in a one-day workshop, their first introduction to ‘facilitative’ supervision concepts, quality improvement, and self-assessment techniques. After the orientation, COPE was introduced in four sites in each country. As is the usual process, district supervisors were encouraged to attend (and mostly they did attend), the first and subsequent COPE exercises, to help site staff address some of the more difficult issues. The district supervisors continued their routine supervisory visits to intervention and control sites during this period.

As per the usual COPE process, external resources were used to conduct short, on-site training in all intervention sites on topics identified as priorities by facility staff. In Kenya, training was organized in facilitative supervision and quality management, immunization updates, and infection prevention. In Guinea, training was organized in facilitative supervision, infection prevention, counseling and information, education, and communication. Any other staff training or updates that occurred during the 15-month period were not considered part of COPE.

Study hypothesis

Evidence from the field of organizational development theory and from participatory action research suggests that individual or team identification and ownership of problems lead to more effective actions than do reporting of issues by external assessors [7–9]. Our hypothesis was that by giving the power of change to facility staff, rather than by imposing change from outside, by giving guidelines about how they might work together rather than what they might work on, and by providing simple tools that reflect quality concerns, staff would start to improve services, and clients would find the facility more attractive and would learn more about how to take care of their children’s health (Figure 1). Consequently, quality improvements in service delivery and client satisfaction with child health services would be more pronounced in settings where COPE was used than in settings where no COPE interventions had occurred.

Figure 1

COPE^® and the quality improvement process.

Study methodology

The study was designed to monitor COPE exercises and associated interventions over a period of about 15 months and then to examine differences in facility staffs’ attitudes and practices, the quality of services, client satisfaction, and services’ utilization between COPE intervention and ‘control’ or non-intervention sites. This report focuses on differences seen between intervention and control sites at the end of the intervention period. Results for Kenya and Guinea are combined; a more complete analysis may be found elsewhere [10].

Eight interventions and eight control or non-intervention sites were included in the study, four of each type in each country. In consultation with Ministry of Health officials, we selected health centers with approximately ten staff, none of whom had undergone IMCI training, but where the Ministry of Health thought such training might potentially happen. Selected facilities were rural or periurban health centers, located many hours’ drive from better-resourced capital cities. Findings from the baseline survey [11] confirmed that intervention and control sites were very similar to key variables: facility size, number of staff, number of child health consultations, range and type of services provided, and the disease pattern of children attending the clinics. The study baseline also confirmed no differences between intervention and comparison sites in service quality, e.g. on indicators such as availability of equipment and supplies for child health service provision and infection prevention, lack of information and outreach to clients, adequate infrastructure, and privacy and confidentiality.

External evaluators (one in each country) were hired to collect baseline and end-of-project data and also to note other changes or externalities that might have confounded the study results. Several instruments were used (Table 1) to collect data and allow triangulation of information for data interpretation purposes.

Table 1

Instruments used in the evaluation

Site descriptions (conducted with clinic managers)1
All eight intervention and eight control sites in Kenya and Guinea
Total of 16 sites
Purpose: To have a general picture of the site who works there, what services are offered, how well is it accessed, and what are principle child health problems in the area it serves
Facility audit (conducted with clinic managers)
All eight intervention and eight control sites in Kenya and Guinea
Total 16 sites
Purpose: To measure the availability of child health materials, equipment, and supplies
Interviews with site staff
All eight intervention and eight control sites
Total interviews: 77 in Kenya and 80 in Guinea (with 76 in intervention sites and 81 in control sites)
Purpose: To measure provider knowledge of quality of care issues, feelings about teamwork and personal involvement in problem solving, attitudes to and relationships with clients, attitudes to work, support from management, adequacy of training and information, adequacy of supplies and an enabling working environment, and problems solved and services improvement
Client exit interviews with caregivers of children
All eight intervention and eight control sites
Total interviews: 160 in Kenya and 160 in Guinea (with 10 well children and 10 sick children at each facility, for 80 sick and 80 well in each country)
Purpose: To measure client satisfaction (various indicators), information retained by clients, reporting of quality, client knowledge of medications, management of sick children, immunization timing, and side effects
Observations of caregiver/provider interactions
All eight intervention and eight control sites
Total observations: 160 in Kenya and 160 in Guinea (with 10 sick children and 10 well children at each facility)
Purpose: To measure provider competence (various indicators), information given to clients (medications, management, immunization, and so on), respect for client, privacy, confidentiality, safety issues, and amount of contact time
Focus group discussions with site staff1
All eight intervention sites
Total of eight focus group discussions (one in each intervention site in Kenya and Guinea, with a total of 88 staff)
Purpose: To solicit staff views about the COPE process

¹

All the instruments, above, were used to collect data at baseline and at the end of the evaluation period except these two instruments. The site description was only used at baseline. The focus group discussions were only conducted at the end of the intervention period.

Selection of facility staff and clients to observe and/or interview was purposive. Given the limited number of clients and providers available each day for consultation, that clients were served on a first-in/first-out basis as they arrived at the facility, and in an effort to maintain a normal patient flow, it was neither possible nor desirable to randomly select clients or facility staff. Instead, several provider–client consultations of well and of sick children were observed sequentially each day by evaluators over a period of three to four days, until the desired number of observations were reached. Several women with well and with sick children were interviewed each day, until the desired number of interviews were completed over a three to four day period. Several staff were interviewed each day based on availability and to ensure that different cadres of staff were interviewed, e.g. ensuring that not only doctors and nurses were interviewed but also facility support staff and administrators. At the end of the study period, external evaluators conducted focus group discussions with available staff in intervention sites, ensuring that the group composition represented the different cadres at the end-of-evaluation data collection period. Group discussions were held on the final day of data collection to limit potential bias that could have been created by discussing the value of COPE earlier in the week.

External factors that might have affected the results seen and attributed to COPE were regularly monitored over the intervention period. These factors (such as training events, equipment donations, and facility renovations) were minimal or occurred equally between intervention and control sites.

All quantitative data were entered into an SPSS database, where frequency distributions and cross tabulations were performed. Chi-squared tests were undertaken to examine differences between intervention and control sites.

Results

Range of problems identified and successfully solved by staff in intervention sites

The COPE exercises conducted in all intervention sites yielded remarkably similar problems in each country, ranging from the simplest-to-address, such as cleaning toilets, to more challenging ones, such as reorganizing work space or improving immunization rates.

After 15 months, the great majority of problems listed on facility action plans were considered ‘solved’ by staff. Table 2 summarizes an illustrative list of the wide array of problems identified and solutions that were implemented. Solutions were often creative. For example, in-service training in a variety of areas was identified as a major need, and local solutions involved staff training each other and finding hitherto untapped local resources. In Guinea, solutions to information and service provision gaps went beyond the clinic and into the community, with staff deciding to provide community education on critical child health themes and to organize more service outreach through mobile clinics. Resources to solve infrastructure problems varied, and smaller activities (e.g. minor construction and painting) were financed with cost-sharing money available for use by sites, whereas larger ticket items (e.g. constructing a maternity unit) came from facility staff successfully exerting pressure on district staff to allocate district funding.

Table 2

Illustrative changes in service quality that occurred due to COPE/quality improvement-related efforts at eight intervention sites, as described by focus group discussion participants in Guinea and Kenyan Health Centers [12]

Infrastructure and equipment issues	—Maternity wards built in two centers
	—Most centers improved supply of running water, electricity, or telephones
	—Renovations of some rooms (e.g. waiting rooms)
	—Limited improvements in availability of materials and supplies were noted; newly available items included light microscopes, privacy screens, mattresses and linen, gas lamps
Human resources issues	—Improvements in staff performance improvements included timeliness (in providing services and being on time for work), being better informed, and being better able to perform their duties (due to formal and on-the-job training)
	—Improvements in staff-to-staff relations (between all levels of staff), more open dialogue was occurring between staff
	—Improvements of staff–to–client interactions included staff treating clients with greater respect, ensuring clients had to wait less time for services, had more privacy, were provided more information. Community leaders were more involved in the health centre than before
Changes in service delivery	—Improved record keeping and use of information
	—More systematic outreach services being provided. One example included staff received training and licenses to ride motorcycles to do regular outreach
	—Integration and expansion of the availability of Child Health services occurred—through physical integration of outreach unit services to reorganization of the physical structure and staff responsibilities to ensure the full complement of Child Health services was provided (including family planning and nutrition)

Some issues were never solved during the intervention period. Drug supply, despite renewed advocacy by site staff, remained problematic. Those issues requiring more senior, or off-site administrative approval and direction, such as how to reduce service costs for the poor in Guinea, were not resolved, although not without trying.

Changes in quality of services provided to children and their caregivers

Observations of the client–provider consultation. To view the range of child health services, we included both ‘well’ and ‘sick’ children and their caregivers in the study. Well children were defined as those coming to the facility with a caregiver to be immunized. Sick children were those coming to the facility with a caregiver for treatment of an illness.

Changes in the quality of child health services that were measured are summarized in Table 3. These indicators were selected by using IMCI protocols outlining ideal consultative behaviors of providers and included things like privacy and confidentiality of the consultation, appropriate interpersonal communication, using consultations to identify ‘missed’ opportunities to diagnose other child health problems or provide preventive services, prescribing practices, and ensuring caretakers were sufficiently informed to adequately manage the illness episode at home.

Table 3

Provider performance observations of interactions between providers and child carers in Guinea and Kenya sites at the end of the 15-month intervention period

Observed provider performance	Indicator	% observations of intervention sites	% observations of control sites	Significance (P value)
Sick children and well children (N = 320)
Privacy and confidentiality	Auditory privacy	60.5	40.9	<0.001
	Visual privacy	58.8	40.3	0.001
	Uninterrupted sessions	81.3	57.9	<0.001
	Maintenance of privacy of client records	100.0	94.3	0.002
Interpersonal skills	Offered client a seat	80.5	63.9	0.001
	Greeted client well	63.8	14.4	<0.001
	Explained things well	64.2	15.2	<0.001
	Confirmed client understood	48.1	6.9	<0.001
	Listened to client	75.0	35.6	<0.001
	Was gentle mannered	70.0	23.6	<0.001
	Smiled	55.6	10.0	<0.001
	Maintained eye contact	61.9	20.6	<0.001
Sick children (N = 160)
History taking	Asked about appetite	37.5	17.5	0.005
	Asked about drinking	10.0	0	0.004
	Asked about fever	42.5	25.0	0.02
	Asked about convulsions	13.8	5.0	0.05
	Asked about vomiting	42.5	40.0	0.74
	Asked about cough	45.0	40.0	0.19
	Asked about diarrhea	32.5	26.3	0.38
	Asked about breathing difficulty	20.0	12.5	0.19
Physical examinations	Examined neck for stiffness	45.0	18.8	<0.001
	Examined abdomen	72.5	57.5	0.04
	Checked hands for pallor	26.3	3.8	<0.001
	Took child’s temperature	63.8	38.3	<0.001
	Felt the neck	45.0	18.8	<0.001
	Examined the ears	20.0	10.0	0.07
Prescribing	Correct malaria prescription given	61.9	50.7	0.06
	Correct antibiotic prescription given	78.8	45.9	<0.001
Information given by provider	Continue to breast feed the child	25.0	7.5	0.003
	Give more fluids to child	23.8	5.0	0.001
	Informed at least two aspects of home care	46.3	13.8	<0.001
	Bring child back if no improvement	48.8	12.5	<0.001
	Bring child back if fever continues	35.0	3.8	<0.001
	Bring back if loss of appetite continues	23.8	1.3	<0.001
	Bring child back if vomiting continues	22.5	7.5	0.008
	Bring back if difficulty in breathing	22.5	5.0	0.001
	Bring child back if she/he has convulsions	26.3	5.0	<0.001
	Information on any danger signs	37.5	8.8	<0.001
Well children (N = 160)
Immunization	Discussed immunization schedule	87.5	72.5	0.02
	Mentioned two or more side effects	63.3	25.3	<0.001
	Discussed management of side effects	69.9	36.5	<0.001
Client–provider discussions	Discussed family planning	21.3	2.5	<0.001
	Discussed child’s growth	26.3	16.3	0.03
	Discussed general health	26.3	8.8	<0.001
	Discussed nutrition	42.2	19.6	<0.001

To determine whether there was a similar illness pattern in intervention and control sites, we assessed examined illness symptoms of sick children. In both countries, the observed symptoms were very similar, reassuring us that changes seen between intervention and control sites at end line would not be biased by a different illness pattern and subsequent treatment actions. The most common complaints in children presenting at intervention and control sites, respectively, were fever (85.0 versus 86.3%), cough (56.3 versus 56.3%), diarrhea (25.0 versus 22.5%), and vomiting (18.7 versus 15.0%).

For the well child consultations, indicators were again selected, based broadly on IMCI protocols outlining ideal consultative behaviors of providers, including privacy and confidentiality of the consultation, discussing immunization schedules with caregivers to reduce missed vaccinations, giving information on why the child is being vaccinated, and ensuring informed and adequate home management of side effects of vaccines like fever and discomfort.

After 15 months of the COPE intervention, on almost every indicator of quality of the client–provider observations—whether of well or of sick children—there was a significant difference between intervention and control sites (Table 3). For well children consultations, there were statistically significant differences in all measures of quality. For sick children, most quality indicators also differed, although on a few variables, there was no significant difference in diagnosing or treating sick children at intervention versus control sites. Oral rehydration salts (ORS) were rarely used in either intervention or control sites (5.0 versus 5.0%), despite almost one quarter of all sick children complaining of diarrhea.

Client perceptions of quality of services. Client perceptions of services are important indicators of the quality of services being provided, and although one expects courtesy bias overall, the study sought to determine any differences between clients in intervention and control sites (Table 4). Client perspectives on the way they were treated by staff were solicited by reading several statements and asking clients to say on a Likert (picture) scale, how much they agreed with the statements. For all statements, there was a statistically significant difference between intervention and control sites (P < 0.01).

Table 4

Client ratings and perspectives on services in Guinea and Kenya sites at the end of the 15-month intervention period (N = 320)

Aspect of service	Indicator positive responses	% clients in intervention sites (n = 160)	% clients in control sites (n = 160)	Significance (P value)
Education	Ever heard a health talk at the site	48.8	14.4	<0.001
	Clients know at least 2 aspects of home care	39.2	26.6	0.09
	Clients knows at least 2 ways to know child is deteriorating and to bring the child back	62.0	34.2	<0.001
	Clients know at least 2 immunization side effects	61.9	44.4	0.02
Agreement with statements about staff that day (clients agree strongly)	Provider was very knowledgeable	85.0	60.6	<0.001
	Provider gave enough time for consultation	78.1	34.4	<0.001
	Provider gave privacy	66.9	24.4	<0.001
	Provider gave information about other services	35.0	8.8	<0.001
	Provider explained when to return	64.4	40.6	<0.001
	Provider explained home care	60.0	36.3	<0.001
	Provider generally explained well	81.9	40.0	<0.001
	Provider was respectful	85.0	62.5	<0.001
	Provider listened well	79.4	53.1	<0.001
	Provider was friendly	86.9	56.3	<0.001
	Overall very satisfied with visit	69.8	48.4	0.001
Agreement with statements about services and facility in general (clients agree strongly)	Very good services overall	70.0	39.5	<0.001
	Good links with other services	61.9	31.4	<0.001
	Generally respectful	65.0	24.4	<0.001
	Good confidentiality and privacy	83.0	53.1	<0.001
	Services are safe	73.4	32.5	0.001
	Waiting times are acceptable	32.5	11.9	<0.001
	Services are at convenient times	32.5	15.0	<0.001
	Services have improved in last year	80.0	26.9	<0.001

To measure the effect of the intervention, evaluators also asked clients whether they had observed any changes in service delivery over the past year. Resoundingly, 80% of clients in intervention sites said that they were better than they had been the year before, compared with 26.9% of clients in control sites (P < 0.01).

The role of management, facilitative supervision, and site staff ability to effect change

To measure changes in management and supervision and how this might influence staff behavior, we asked interviewed staff in all sites their opinions about the way the facility was managed and their ability to participate in decision making and influence decisions about service delivery. In all management and self-efficacy areas that were covered during interviews (being valued as a staff person, management taking an interest, being able to make suggestions, feeling part of a team, feeling that management is responsive, and feeling that colleagues have high morale), the staff in intervention sites had significantly higher morale and were more satisfied with their jobs than at comparison sites (Table 5).

Table 5

Provider perspectives on management and supervision in Guinea and Kenya sites at the end of the 15-month intervention period (N = 157)

Perspectives	Indicator positive responses	% staff in intervention sites (n = 76)	% staff in control sites (n = 81)	Significance (P value)
Facility management	Management more responsive lately	69.7	23.5	<0.001
	Management encourages training	67.1	27.2	<0.001
	Management discuss service statistics	42.1	27.5	0.19
	Management likes to solve problems	64.5	25.9	<0.001
	Management takes an interest in my job	80.3	56.8	0.007
	Staff morale is high	82.9	36.3	<0.001
	I feel part of a team	72.4	45.0	0.005
	My opinion is valued by management	65.8	24.7	<0.001
Outside supervisors	Supervisors include us in their discussions	63.2	32.1	<0.001
	Supervisors help us solve problems	61.8	21.0	<0.001
	Supervisors help us with training	65.8	22.2	<0.001
	Supervisors help us with supplies	65.8	18.5	<0.001
	We truly benefit from supervision	60.5	22.5	<0.001

Facilitative supervision is promoted by COPE to reinforce site self-improvement efforts and staff self-efficacy, and off-site/district supervisors received training in facilitative supervision early in the project. Staff who were interviewed were asked their perceptions of management support, morale, and their opinions about supervision by off-site/district supervisors. Although there was no significant difference in the number of recent supervisory visits between intervention and control sites, the responses to statements about what supervisors did during their visits (working with providers, helping with problem solving, helping with training, and helping with supplies) were significantly (P < 0.01) different between the two types of sites (Table 5).

Discussion

Significant improvements seen in quality of services

On almost every quality indicator, whether reported by staff, observed by evaluators, or reported by clients, the intervention sites performed statistically significantly better than the control sites only 15 months after these low-key interventions began. Although there were (expected) quality improvements in both countries related to the more direct contributions made by the project per se, such as improvements in infection prevention after training, there was also evidence of a whole range of other improvements that resulted from staff actions themselves. In the intervention sites, we observed greater availability of services being provided in cleaner, more pleasant, more private settings. We also observed (confirmed by clients) more respect and information for clients, more privacy, with improved provider interpersonal communication skills, use of improved diagnostic skills, improved home care instructions, some improvement in prescribing practices, and improved immunization practices. We also found more informed and more satisfied clients, and their acknowledgment that changes in services had occurred over the past year.

Why did COPE trigger staff actions to improve quality?

The COPE exercises only suggest what standards of care might be; there are no specific interventions. Limited short training requested by staff in information, education, and communication approaches, infection prevention, and facilitative supervision was conducted, but the changes seen in this study are much broader in scope and begin to address the underpinnings of quality services. Nobody told staff that they needed to treat clients better, give out more information, ensure uninterrupted consultations, and take better histories. Working through the COPE exercises enabled those individuals willing to look critically at themselves to plan and make changes to self-identified problems. Working through the exercises as a group of staff helped foster a critical mass of enabled workers.

With an open-ended intervention like COPE, what led staff to take specific and sometimes bold actions to improve quality of services? Staff generally know what needs to be done to provide quality services. But they sometimes forget; or they are unable to do a good job because they lack the tools or the technical expertise, or they lack feedback on their performance; or they are so demoralized that they have given up trying to understand and interact personally with their clients. We had hypothesized that the COPE intervention would lead to personal and organizational change that providers would feel empowered, more confident and free to act, assume ownership of the problems (and the solutions), have raised morale and commitment, be more reflective, and feel better supported. Findings from end-of-evaluation staff focus group discussions, reported elsewhere [12], confirmed that staff did indeed feel that they had begun to break down some of the communication barriers and inertia running through their health services and that COPE had helped to provide the fertile ground upon which organizational change could occur, changes that led to improved quality of service and enhanced client satisfaction. Staff told us that the fact of outsiders not identifying the problems, not suggesting the answers, and not providing the solutions, but instead creating an enabling environment for staff to do those things themselves, is what stimulated action and created change. This very ownership of problems and their solutions, although daunting at first, seems to have had a strong impact on staff attitudes toward their work environment and in changing their own behaviors/interpersonal interactions with other site staff as well as with clients. This was reinforced by feelings that management, supervisors, and clients appreciated them and were relying on them to make good decisions.

What types of issues were not affected by COPE?

Although COPE could effect changes on service quality in many areas, there were a few indicators where there was little or no observable difference between intervention and control sites. For example, there were generally poor prescribing practices in both intervention and control sites in both countries. Although COPE can raise issues such as these, some problem areas still will require specific technical skills and knowledge to address them.

There are other areas where staff are constrained in their ability to take action. The data showed that there were little observable or sustainable differences between the intervention and control sites in availability of drugs and equipment, even though many intervention sites had taken steps to work with the local health committees to make funds available from the community coffers for such purchases. The important role of external support from district management committees, supervisors, and community health councils is crucial to solve such problems and to keep facility staff engaged in their own problem resolution efforts.

Conclusion

The IMCI approach to child health has been shown to be useful in helping health care providers recognize the need for holistic curative and preventative care for children and in training providers to better manage childhood illness. Efforts, though, are constantly thwarted by unmotivated and unsupportive colleagues, inadequate facility infrastructure, generally poor quality of care by other staff (poor information, counseling, and attention to privacy), non-facilitative supervision, failing health service support systems (supplies, record keeping, and effective infection prevention), and poor community linkages.

COPE is a very low-tech, easy to do quality improvement process, and our study confirms that it can have a very dramatic effect on the way people work and the services they provide. Providers feel better about their work, and the clients feel better about the services. This project achieved significant results after one person visited each site four times over a 15-month period, encouraged supervisors to participate, and provided a very small amount of money for (mostly on-site) training in site-determined subjects.

This study demonstrated that the COPE quality improvement initiative can in fact bring facility staff, managers, and supervisors together to support each other, can raise staff morale, can effect changes to the facility infrastructure and functioning, can improve provider attention to the care of sick children and immunizations, and can help providers share information between themselves.

Given political commitment, scaling up this quality improvement process in any country would not be a costly or difficult exercise. In addition, we suggest that COPE processes could be implemented in a way to complement other training and quality improvement efforts, as its self-assessment philosophy and tools create a more enabling environment for change, and hence, could make other quality improvement efforts more effective. To this end, more research could be done to evaluate how COPE might be undertaken in conjunction with IMCI training, for example, to accomplish improvements in services that can help lead to better health for children.

¹

COPE is a registered trademark of the U.S. Patent Office.

Author notes

¹EngenderHealth, New York, USA, and ²Consultant, EngenderHealth, New York, USA