Improving Compliance with Standards for Essential ...pdf.usaid.gov/pdf_docs/PNACQ767.pdf · cases, more than one QA ... Improving Compliance with Standards for Essential Obstetric

Q U A L I T Y

A S S U R A N C E

P R O J E C T

Center for Human Services • 7200 Wisconsin Avenue, Suite 600 • Bethesda, MD 20814-4811 • USA

Q U A L I T Y I M P R O V E M E N T C A S E S T U D Y

Improving Compliance withStandards for Essential

Obstetric Care in Bolivia

J une 2 0 0 2

QUAL IT YASSURANCE

PROJECT

TEL ( 301 ) 654 - 8338

FAX ( 301 ) 941 - 8427

w w w. q a p r o j e c t . o r g

The Quality Assurance (QA) Project is funded by the U.S.Agency for International Development (USAID), under ContractNumber HRN-C-00-96-90013. The QA Project serves countrieseligible for USAID assistance, USAID Missions and Bureaus, andother agencies and nongovernmental organizations thatcooperate with USAID. The QA Project team consists of primecontractor Center for Human Services; Joint CommissionResources, Inc.; and Johns Hopkins University (including theSchool of Hygiene and Public Health [JHSPH], the Center forCommunication Programs [CCP], and the Johns Hopkins Programfor International Education in Reproductive Health [JHPIEGO]).The QA Project provides comprehensive, leading-edge technicalexpertise in the design, management, and implementation ofquality assurance programs in developing countries. Center forHuman Services, the nonprofit affiliate of University ResearchCo., LLC, provides technical assistance and research for thedesign, management, improvement, and monitoring of healthsystems and service delivery in over 30 countries.

About this series

The Case Study series presents real applications ofquality assurance (QA) methods in developing countries atvarious health system levels, from national to community.The series focuses on QA applications in child survival,maternal and reproductive health, and infectious diseases.Each case study focuses on a major QA activity area, suchas quality design, quality improvement, communication anddevelopment of standards, or quality assessment. In somecases, more than one QA activity is presented.

Quality improvement is a systematic process of addressingthe gaps between current practices and desired standards.Effective approaches to quality improvement includeindividual problem solving, rapid team problem solving,systematic team problem solving, and process improve-ment. These methods vary in the time and resourcesrequired and the number of people who participate.Regardless of the rigor and intensity of the method used,quality improvement approaches usually share four basicsteps: identification of opportunity for quality improvement,analysis of improvement area, development of possibleinterventions to address a need for improvement, andtesting and implementing interventions. Sometimes, whenthe potential solutions to a problem are clearly defined, ashorter quality improvement activity focused on testing thealternatives is used.

This case study illustrates how the quality improvementmethodology—specifically rapid team problem solving—can be used to target areas for improvement and producerapid, yet dramatic, improvements.

Improving Compliance with Standards for Essential Obstetric Care in Bolivia

AcknowledgementsThis case study is based on the outstanding work of facilita-tors at the Santa Cruz Health Department (SEDES) andRosmery Chavez, field coordinator for the Latin Americanand Caribbean Regional Initiative to Reduce MaternalMortality (LAMM). They provided extensive technicalassistance to the Ichilo, Bolivia, quality improvement teamswhose work is described here. The work involved manyindividuals at the Yapacani, San Carlos, and Buena Vistahospitals, individuals who share the goal of providing thehighest possible quality of care. The facilitators who contrib-uted to the success of these teams include Dafne Delgadillo,Oscar Moreno, and Ruth Galvez. SEDES, in particular CletoCasseres and Federico Urquizo, and Ichilo district person-nel, especially Abel Monasterios and Sandra Fernandez,also supported the teams. Jorge Hermida, Associate QAProject Director for Latin America, provided technicalassistance and encouragement to the teams.

Diana Silimperi, Jorge Hermida, Stephane Legros, Ya-ShinLin, and Lani Marquez provided generous contributions inthe development of this case study. A special thanks to DavidNicholas (QA Project Director), Annette Bongiovanni (USAID,LAC Bureau), and Jim Heiby (USAID, Global Bureau) for theirsupport and direction in LAMM and QA Project activities. Wealso thank Beth Goodrich for editorial review and SuchaSnidvongs for graphic design.

Recommended citationAskov K and Chavez R. 2001. Improving Compliance withStandards for Essential Obstetric Care in Bolivia. QualityAssurance Project Case Study. Published for the U.S. Agencyfor International Development (USAID) by the QualityAssurance Project: Bethesda, Maryland, U.S.A.

Improving Compliance with Standards for Essential Obstetric Care in Bolivia 1

Q U A L I T Y I M P R O V M E N T C A S E S T U DY

Improving Compliance with Standardsfor Essential Obstetric Care in Bolivia

Background

An estimated 370 women diein Bolivia from pregnancy-related causes for every100,000 live births.a Thismaternal mortality rate is oneof the highest in LatinAmerica and more thandouble the regional average of 162 maternal deaths per100,000 births. Maternal deaths stem not only from a lackof knowledge about obstetric danger signs and complica-tions, and difficult accessibility to essential obstetric care(EOC),1 but also from the inability of healthcare providers todetect and manage obstetric complications.

The Quality Assurance (QA) Project partners with the PanAmerican Health Organization (PAHO) in the Latin Ameri-can and Caribbean Regional Initiative to Reduce MaternalMortality (LAMM) to address knowledge of obstetric danger

1 EOC includes the detection and management of problem pregnancies (e.g.,severe anemia, diabetes, etc.), medical treatment of complications(hemorrhage, sepsis, shock, eclampsia, anemia, etc.), monitoring labor(including partograph), neonatal special care, manual procedures (e.g.,removal of placenta, episiotomy repair, etc.), surgical interventions,anesthesia, and blood replacement.

Improving Compliance with Standards for Essential Obstetric Care in Bolivia2

Abbreviations

EOC Essential Obstetric Care

JSI John Snow, Inc.

LAMM Latin American and Caribbean Regional Initiative to ReduceMaternal Mortality

MOH Ministry of Health

PAHO Pan American Health Organization

QA Quality assurance

QI Quality improvement

SEDES Servicio Departamental de Salud—Santa Cruz (Santa CruzHealth Department)

UNICEF United Nations International Children’s Emergency Fund

USAID United States Agency for International Development

signs and access to EOC at the community level, while alsoimproving the ability to detect and manage obstetriccomplications in health facilities. Funded by the UnitedStates Agency for International Development (USAID),LAMM has provided technical assistance to the Ministriesof Health (MOH) of Bolivia, Ecuador, and Honduras since1997 with the dual objective of reducing national levels ofmaternal mortality and developing a model—to be repli-cated in the Latin America and Caribbean region—toreduce maternal mortality.

The QA Project initiated technical assistance to threehospitals and one health center in the district of Ichilo,Bolivia, in 2000 with the objective of improving the ability ofhealthcare providers and facilities to detect and manageobstetric complications. One way to improve the quality ofEOC is to initiate quality improvement (QI) teams in eachhealth facility to improve prenatal, labor, and delivery care,important components of EOC. Six multidisciplinary QIteams (Table 1) were formed in August of 2000 after aworkshop that introduced participants to the concepts ofquality assurance. This case study highlights the experi-


ences and results of the labor and delivery teams at thehospitals.2

Table 1. Quality Improvement Teams (Ichilo,Bolivia)

Hospital Component of EOC for Improvement

Yapacani Hospital / Team 1 Labor and delivery

Buena Vista Hospital / Team 1 Prenatal care

Buena Vista Hospital / Team 2 Labor and delivery

San Carlos Hospital / Team 1 Prenatal care

San Carlos Hospital / Team 2 Labor and delivery

Jampicuna Health Center / Team 1 Prenatal care

With the specific objective of improving the quality ofmaternal care, teams focused on improving processeseither for prenatal care or for labor and delivery. Theseteams generally included physicians, nurses, auxiliarynurses, laboratory technicians, statisticians, administrators,and when possible, an obstetrician and pediatrician.3

Guidance was provided by several individuals. Thehospital directors frequently participated in QI activities andclosely followed the teams’ achievements. In addition, alocal QA Project field coordinator, trained by the QA Projectto provide technical assistance during these qualityassurance activities, provided continued support to theteams. Lastly, four facilitators from the Santa Cruz HealthDepartment (SEDES) provided guidance. The facilitatorswere physicians or nurses and were selected for their

2 The main differences between the health center and the hospitals is thathospitals are open 24 hours per day, are equipped to provide emergencyobstetrical care, and are covered by social security. Because of thesecharacteristics each hospital delivers more babies than the health center.

3 San Carlos hospital is the only facility that, at this writing, has anobstetrician and pediatrician, as it is a referral hospital for obstetriccomplications.


technical knowledge and leadership skills. The LAMM fieldcoordinator had trained the facilitators in the application ofthe QI methodology, which was new to the Ichilo district,and mentored the facilitators during the QA activities.

4 Bolivia is divided into nine departments consisting of districts andmunicipalities.

It is important to note thatthese QI efforts took place inthe midst of health sectorreform. The national healthsystem has been highlydecentralized since 1994when the Popular Participa-tion Act transferred responsi-bility for infrastructure,

financial resources, maintenance, and administration tomunicipalities.b Next, the Administrative Decentralization Acttransferred the administration of human resources to eachdepartment.4 Bolivia then sought to improve access tohealthcare for pregnant women and children under the ageof five by creating a national insurance system. Since 1996this system has provided universal coverage for all pregnantwomen and all children under the age of five. It coversambulatory care, medical assistance, medical attention forpregnancy, delivery, postpartum care, obstetric emergen-cies, pharmaceuticals, and hospitalization for acutediarrhea and pneumonia.

The success of the QI methodology depends on the activeparticipation of personnel and community members tocreate change to improve the quality of care and to mobilizelocal resources to implement those changes. Bolivia’shighly decentralized healthcare system—particularly formaternal and child health—created a responsive environ-ment for quality improvement in EOC services. In general,Ichilo’s QI teams received only technical assistance fromthe QA Project; the improvements they made were usuallylow cost or no cost. This aspect of the QA Project’s technicalassistance contributes to the low cost of this model and thesustainability of quality assurance activities.


Quality Improvement Methods

The Ichilo QI teams had the specific objective of monitoringand improving compliance with clinical and administrativestandards.5 The emphasis on compliance with standardsaimed to reinforce the EOC skills training that LAMM hadprovided to nearly every healthcare provider in the district.Each QI team began by agreeing to focus on the standards

in Table 2.

The partograph is a graphic instrument used byhealthcare providers to monitor progress in labor andto assess and record the condition of the mother andfetus. It is effective for early detection of prolongedlabor.

These laboratory tests are critical to the quality ofcare during labor and delivery. The testing ofhemoglobin can detect anemia; tests for the bloodtype and Rh factor are necessary in case a bloodtransfusion is needed; and tests for syphilis are criticalto the health of the mother and newborn.

Some of the most serious and life-threateningcomplications, such as hemorrhage or sepsis, occurduring the postpartum period of 2 hours afterdelivery. In order to detect complications in a timelymanner, it is important to standardize what must bemonitored and when.

A protocol of care is a plan, or set of steps to follow,in the management of a specific patient.c It is criticalfor documenting the care provided to each patientand planning for any follow-up procedures or actions(e.g., how often to assess the patient, what toassess, what treatments or procedures may benecessary).

All women in labor and delivery willbe monitored with a partograph.

All women in labor and delivery willhave laboratory exams for hemoglo-bin, blood type, Rh factor, and syphilis.

Every woman will be examined every30 minutes for the first 2 hours afterdelivery for vital signs, bleeding, anduterine status.

Table 2. Standards of Care for Labor and Delivery (All Laborand Delivery Teams, Ichilo)

Standard Significance of the Standard

All women in delivery will have a pro-tocol of care clearly documented.

5 A standard is an expectation of quality that is explicit (written) or implicit(understood). Explicit standards appear in a variety of forms (e.g.,specifications, protocols, procedures) and may be developed by a numberof different organizations such as Ministries of Health, professionalorganizations, accrediting organizations, or by a team of workers in ahealthcare system. See also Ashton 2001 in the endnotes.


Problems (also called “opportunities for improvement”)differ in a number of aspects, such as the complexity of theproblem, the amount of existing data available on theproblem, and the level of insight and knowledge teammembers have about the problem.d The QA Project re-sponds to this variation by identifying four different ap-proaches to improving quality. Each approach (individualproblem solving, rapid team problem solving, systematicteam problem solving, and process improvement) followsthe steps in Figure 1.

Figure 1. Steps to Quality Improvement

6

The LAMM field coordinator recommended rapid teamproblem solving for the Ichilo teams. In this approach, a QIteam studies a healthcare system and then develops, tests,and implements a series of small, incremental changes tothe system or its components. This approach is “rapid”because it is used when team members have insight intothe causes of and potential solutions to the problem.(Having these insights minimizes the need for data collec-tion to determine the root cause[s] of the problem.) Thisapproach was appropriate for the teams in Ichilo as theteam members had insights into the potential causes of the

Plan

Act

Study

1. Identify the problem

2. Analyze the problem

3. Develop solutions

4. Test and implement solutions

Do


problems selected and wanted toachieve rapid, initial results. In fact,the Ichilo teams began to achieveresults within a month or two. It isimportant to note, however, that theteams could successfully applyrapid team problem solving becausethey received ongoing support andguidance from the facilitators, MOHdistrict and provincial supervisors,and the QA Project.

After each team identified the opportunities for improve-ment it would address, the facilitators helped them decidewhich opportunities could be addressed by managementaction and which were sufficiently complex to require teamproblem solving (see “Selecting Solutions” below in Step3). This division of work allowed the teams to focus on themost complex issues. Management action generallyyielded dramatic and sustained results; however, whenteams were not satisfied with the results from managementaction, they revisited the problem and addressed it usingrapid team problem solving.

Step One: Identify the Problem

Problems, or opportunities for improvement, becomeapparent in several different ways. For instance, anadverse event may occur, a client or someone on the staffmay notice that something should be better, or an externalevaluation may uncover a problem that staff had over-looked or felt unable to correct. In Ichilo, opportunities forimprovement were identified in the context of poor compli-ance with certain standards. The hospitals intentionallyassessed quality by identifying standards staff thoughtshould be improved, measuring compliance with thestandards, and assessing whether compliance wasacceptable. Therefore, the teams in Ichilo identifiedopportunities for improvement by selecting standards,developing key indicators, and setting baselines for thoseindicators.


6 The Bolivian MOH developed and published these standards in “Atención ala Mujer y Recién Nacido” with the technical assistance of MotherCare(JSI, UNICEF, and PAHO).

8

Selecting key standards. The QI teams reviewed BolivianMOH standards for EOC6 and selected key technical standards(e.g., postpartum monitoring and use of the partograph) andadministrative standards (e.g., the accurate completion ofclinical records).

The teams also developed standards that had not been promul-gated by the MOH. They identified what they believed were themost important quality issues for their communities and devel-oped standards for those issues. For example, the Buena Vistateam felt that personalizing the care of women during labor anddelivery was important to their community, so they developedstandards to give each woman the options of choosing herposition for delivery and permitting a family member to accom-pany her during delivery. Developing and implementingstandards outside the MOH’s purview demonstrated the teams’ability to improve services from the clients’ perspective.

Although each team worked independently, all three labor anddelivery teams selected the four standards in Table 2 from theMOH publication.e These four standards comprise most of thediscussion in this case study; some descriptions of the stan-dards each team conceived to meet community expectationsare also included.

Developing and measuring key indicators. The teamsdeveloped indicators to measure the level of compliance foreach of the selected standards. Next, each team determinedthe baseline of each indicator. To determine the baseline, oneor two team members and the QA Project facilitator reviewedthe medical records for all women who delivered at the hospitalover the previous three months. This review took approximatelyone month and was completed in stages. Yapacani Hospitalbegan first and finished in August 2000. Buena Vista and SanCarlos Hospitals began next and finished in October 2000. Thiseffort provided information about the actual level of compliancewith each standard and highlighted the difference betweenactual and perfect compliance (Table 3).


Table 3. Standards, Indicators, and Baseline Data (All Laborand Delivery Teams)

All women in labor anddelivery will be monitoredwith a partograph.

All women in labor anddelivery will have laboratoryexams for hemoglobin,blood type, Rh factor, andsyphilis.

Every woman will beexamined every 30 minutesfor the first 2 hours afterdelivery for vital signs,bleeding, and uterinestatus.

All women in delivery willhave a protocol of care thatis clearly documented.

Standard IndicatorSan CarlosN=52Oct. 2000

YapacaniN=124Aug. 2000

Buena VistaN=49Oct. 2000

Percentage of deliveriescorrectly monitored witha partograph

Percentage of attendeddeliveries with laboratoryexams for hemoglobin,blood type, Rh factor, andsyphilis

Percentage of womenmonitored for vital signs,bleeding, and uterinestatus every 30 minutesfor the first 2 hours afterdelivery

Percentage of women whodeliver and have aprotocol of care

0%

33%

0%

8%

9

24%

83%

24%

83%

12%

7%

16%

77%

An indicator is satisfied only if compliance is complete. Forexample, the indicator for laboratory testing would not bemet if the healthcare provider tested for blood type but notsyphilis. Likewise, the postpartum monitoring standard issatisfied only if the provider completed and recorded theentire regimen every 30 minutes for the first two hours afterdelivery.

Defining the problems. The baseline data provided thenecessary information to define compliance problems.Examples of these definitions (in terms of compliance witheach standard) are in Table 4.


76% of women in labor and delivery are notmonitored with a partograph.

93% of women in labor do not have laboratoryexams for hemoglobin, blood type, Rh group, andsyphilis.

100% of women in immediate postpartum careare not strictly monitored for vital signs, bleeding,and uterine status every 30 minutes for the first 2hours after delivery.

Buena Vista Hospital

Yapacani Hospital

San Carlos Hospital

10

Table 4. Examples of Definitions of IdentifiedProblems

Selecting an approach to quality improvement.Because the teams elected to address all the problemsidentified, the QA Project facilitator helped them to deter-mine which could probably be addressed by managementaction and which warranted team problem solving. Whenteams elected management action, the hospital directortook responsibility for improving compliance. TheYapacani and San Carlos teams chose managementaction for the partograph-monitoring standard, and theBuena Vista team chose this method for postpartummonitoring.

Assuming Ownership of the Results

Initially, the team members weredisappointed with their baselinemeasurements, and some of themblamed others. The team facilitatorsmet with each team to discuss the factthat everyone was responsible for theresults and that if they worked togetherto apply the QI methodology, theseproblems could be overcome. Based onthis discussion, team membersassumed ownership of the baselinemeasurements and elected to addressall of the problem areas.

The facilitator ex-plained to the teamsthat if managementaction did not solve aproblem, rapid teamproblem solving wouldbe necessary. Indeed,some teams did findthat managementaction did not satisfac-torily solve a problem,so they revisited it.(Details are in “Act onthe Results” below.)


Building Multidisciplinary Teams

When the labor and delivery teams atYapacani and Buena Vista drew a high-level flowchart of the process for takinglaboratory exams of women in labor, theyrealized that the laboratory technicianscould help them understand the processand propose potential solutions. Theteams invited these technicians toparticipate in the QI teamwork. Thetechnicians became active team membersand also contributed to the developmentand implementation of solutions in otherareas. For example, a Yapacani techniciannow conducts monthly client satisfactionsurveys to assess the information patientsreceive about respiratory breathing andpositioning during labor and delivery.

11

Step Two: Analyze the Problem

During the “analyze” step, a team considers what it needsto know or understand about a problem before proposingsolutions. During this step, team members ask: Who isinvolved or affected by the problem? Where and when doesthe problem occur? What happens when the problemoccurs? Why does the problem occur? Rapid team problemsolving emphasizes using existing data and brainstormingamong team members to understand the problem. Theteams drew on their collective knowledge of the labor anddelivery processes to draw flowcharts and brainstormabout possible problem causes. The following examplefrom Yapacani exemplifies how the teams performedproblem analysis.

Analyzing processes. A high-level flowchart shows themajor steps in a process, providing a basic picture of theprocess and indicating where each member of thehealthcare team participates. It helps people understandthe processes within a system. By drawing high-levelflowcharts of each system that ended in an undesirableresult, the teams couldsee where theproblems originated.

For example, to fullyunderstand theprocess for immediatepostpartum care as itwas being performedat Yapacani, the laborand delivery teamdrew the flowchart inFigure 2. As notedabove, this standardassures the timelydetection of anypostpartum complica-tions by monitoring, forthe first two hours afterdelivery, for vital signs,


Thenursingshift ends

Activitiesarereported

Medicalorders arecarried out

Patientleaves thedeliveryroom

Patient isadmitted to thepostpartumroom

Hypothesizing about causes of the problem. A cause-and-effect analysis helps teams see links between problemsand their potential causes. The fishbone diagram is one type ofcause-and-effect analysis; it helps teams to brainstorm aboutpossible causes of a problem and to group those causes intocategories. The Yapacani team developed the fishbonediagram in Figure 3 to understand and hypothesize (i.e., maketentative assumptions about) the underlying causes contribut-ing to the lack of monitoring during immediate postpartumcare. That is, the team hypothesized causes that would answerthe question, Why does monitoring stop when the nursing shiftends? The hypothesized causes were grouped into threecategories: procedures, personnel, and methods. The lack ofstandardization in monitoring and reporting key informationabout postpartum patients cut across all three categories (seebold text in Figure 3). It is important to note that although thisdiagram helped the team members collect and organize theircollective insights about potential causes of the problem, it didnot prove or single out a primary cause: certainty of causesrequires quantitative data.

12

uterine status, and bleeding. The team needed to under-stand why this monitoring was not occurring. In discussingthe process for this care and devising the flowchart, theteam members noticed a lack of continuity of care betweennursing shifts. The flowchart showed the team that theexisting process for immediate postpartum care did nothave a system to assure continuous monitoring of patientsacross two nurses’ shifts.

Figure 2. Flowchart of the Monitoring Processduring Immediate Postpartum Care (Labor andDelivery Team, Yapacani Hospital)


METHODS

PROCEDURES Lack of time

No reportingmechanism

No one is responsiblefor this action

Lack of internalorganization/feedback

No standardmethod formonitoring inpostpartum

100% of women in labor arenot monitored for vital signs,uterine status, and bleedingevery 30 minutes during thefirst 2 hours after delivery.

Figure 3. Fishbone Diagram of HypothesizedCauses for Lack of Monitoring during ImmediatePostpartum Care (Yapacani Hospital)

PERSONNEL

13

Analysis tools like the flowchart and fishbone diagramhelped the teams to understand the processes andpossible causes for the identified problems.

Step Three: Develop Solutions

The third step, “develop solutions,” uses information fromthe previous step to formulate hypotheses about whatsolutions would likely reduce or eliminate the problem andin turn improve the quality of care. Having identified andanalyzed possible causes of the problems, the teams hadample information for developing hypotheses aboutsolutions. Reflecting on the causes they had hypothesizedin Step 2, they tried to identify solutions that might be linkedto those causes. They hypothesized about what wouldreduce the problem or make it stop occurring: a proposedsolution. Proposed solutions remain theoretical until theyhave been tested and have demonstrated improvement (asshown in the next step). Since rapid team problem solvingsupports the development of many small, incrementalsolutions that the team develops to gradually reduce aproblem, it was anticipated that more than one solution


might be proposed for each problem and that solutionsmight be implemented one at a time, with each tested tomeasure its impact.

Brainstorming to develop solutions. Using theinformation related to the causes of each problem, theteams generated a list of solutions for each problem usingbrainstorming. Brainstorming is an activity where a group ofpeople (individuals can do it, but having more peoplegenerates more ideas) raise ideas without evaluatingwhether the ideas are good or not. This process taps intogroup knowledge and creativity in a forum where ideas arefreely generated. Once several ideas have been expressedand recorded, the group discusses them, elaborates onthem, combines them, and prioritizes them.

Selecting solutions. After the teams had developed alist of potential solutions, the facilitator and team discussedthe advantages, disadvantages, and feasibility of each.This discussion helped the teams to select solutions thatbalanced feasibility and the likelihood of reducing theproblem. For example, the team at Buena Vista hypoth-esized that one reason why women in labor did not receivelaboratory exams was because the laboratory technicianwas off duty (e.g., holidays and weekends). The teammembers generated a number of solutions to address thisissue, such as calling the technician to the hospital when atest was needed, hiring another technician, and trainingnurses to conduct these tests. After discussing the advan-tages and disadvantages of each proposed solution, theteam selected7 training nurses as the most viable, sincecalling in the off-duty technician was not favorable andbudget restrictions precluded hiring another. Table 5summarizes the information gathered thus far in the QIprocess.

Table 5 shows that although the labor and monitoringteams addressed many of the same problems, many

7 Various prioritizing tools, such as voting, may be used to reach agreementon which proposed solution will be tested first; see Massoud et al. 2001(endnotes), p. 54-58.


2.DO the test.

3.STUDY the resultsachieved as comparedto the desired results.

1.PLAN the test.

4.ACT on the results, modifyingor creating a new solution if theimprovement is insufficient.Implement the solution ifimprovement is sufficient.

Step Four: Test and Implement Solutions

This final QI step tests the proposed solutions to determinewhether they yield sufficient improvement. In rapid teamproblem solving, a number of solutions can be incremen-tally developed, tested, and, if proven effective, imple-mented into the system. Solutions are tested using the Plan,Do, Study, Act Cycle (PDSA or Shewhart’s Cycle for Learn-ing and Improvement)f shown in Figure 4.

Figure 4. PDSA Cycle to Test Solutions

Planning the test. At this stage the QI teams had acomprehensive list of possible solutions and had toorganize all the activities and resources that would have tobe in place to test these solutions. Each team created atable that identified: (a) the activities, (b) the human andmaterial resources needed for completing the activities, (c)the individual(s) responsible for performing the activities,and (d) the timetable for performing and completingactivities. The tables served as graphic displays for theirplan for testing solutions (Table 6 is an example). Even aftersuccessfully carrying out the plan, the teams continue tomonitor compliance with standards, develop solutions to

different solutions were proposed. This variety helpedteams learn from each other’s experiences as well asadapt and implement each other’s solutions.

Improving Com

pliance with Standards for Essential O

bstetric Care in Bolivia

16

No system exists for requesting laboratory testing supplies.

Identified Problem Method Hypothesized Cause(s) Proposed Solutions

Yapacani Hospital

Laboratory exams Team problem solving

Table 5. Information Gathered in Steps 1 through 3 (All Labor and Delivery Teams)

Exams are not done on weekends and holidays when thelaboratory technician is off duty.

Postpartum monitoring No systematic way exists to monitor key areas inimmediate postpartum.

Create a rotation staff schedule so that staff are alwayspresent to monitor key areas.

Staff do not know how and when to monitorpostpartum patients.

Train staff in monitoring patients.Create a reporting system to include immediate postpartummonitoring in the clinical record.

Protocols Management action Communication is lacking about management’sexpectations for staff in complying with the protocols.

Communicate the baseline results to all personnel andoutline expectations for improvement.

Provide practical and theoretical training inlaboratory testing.

Personnel do not know how to conduct thelaboratory exams.

Team problem solving

Create a system to request laboratorysupplies every three months.

Improving Com

pliance with Standards for Essential O

bstetric Care in Bolivia

San Carlos Hospital

Laboratory testing Team problem solving Personnel require reinforcement and practice inconducting laboratory exams.

Inform all personnel that these laboratory tests for womenin labor and delivery are a standard of care.Present a half-day practical training in conductinglaboratory exams.


Monitoring laborwith a partograph

Management action Communication is lacking about management’sexpectations for staff in complying with the protocols.


Protocols Management action Communication is lacking about management’sexpectations for staff in complying with theprotocols.

Postpartum monitoring Management action Communication is lacking about management’s expec-tations for staff in complying with the protocols.


Managementaction

Communication is lacking about management’s expec-tations for staff in complying with the protocols.


Buena Vista Hospital

Monitoring laborwith a partograph

Team problem solving Personnel require reinforcement and practice in theirskills to monitor labor with a partograph.

Present a one-day reinforcement course for all nurses andphysicians.

Identified Problem Method Hypothesized Cause(s) Proposed Solutions

Protocols

1

7


Table 6. A Sample Plan to Test Solutions (Laborand Delivery QI Team, San Carlos Hospital)

X

Timeline(Weeks)

Activity HumanResources

MaterialResources

ResponsibleIndividual(s)

Conduct a half-day training inthe use of thepartograph.

General physiciansObstetricianNurses

Partographs Generalphysicians

ObstetricianFacilitator

1 2 3 4 5 6

Create a rotatingschedule of staff forimmediate postpartummonitoring.

ObstetricianNurses

Suppliesnecessary toconduct themonitoring

Head of nursingSupervisor of nursing

Create a reportingsheet that will beincluded in themedical record.

Nurses Reporting sheets Head of nursingSupervisor of nursing

Request a bloodpressure gauge.

District director Paper for theletter

Hospital director

X

X

X

18

Doing the test. The teams then carried out the tests,using their plans as guides. The tables helped the teamsremember who was responsible for each activity and tomaintain accountability for its implementation. If the plansor activities changed, the teams documented the changesso that successful interventions could be preciselyreplicated and/or expanded and ineffective changes couldbe modified in the implementation of the intervention.

Studying the results. Allowing time for change to occuris important because the interventions—even if effective—may not show immediate results. Each month for severalmonths, the QA Project facilitator and a team memberwould review all medical records related to deliveries tocompile data for each indicator. Changes in the indicatorsshowed the effect of each set of interventions. At Yapacani,an average of 40 (ranging from 33 to 48) records wasreviewed each month from August 2000 to February 2001;at Buena Vista, an average of 15 (ranging from 10 to 19)

yield further improvements, and incrementally test andimplement those solutions.


records was reviewed each month from October 2000 toFebruary 2001; and at San Carlos, an average of 12 (from8 to 23) records was reviewed from October 2000 toFebruary 2001. The monitoring of the indicators wouldreveal which interventions were successful and whichrequired further modification.

Acting on the results. The teams met monthly to reviewthe results as the activities continued and increasinglycaused improvements. Most of the interventions yielded somuch improvement that the teams decided to implementthem permanently (see “Results”). In addition, some werealso implemented in other hospitals, and some wereadopted at the district level.

Figure 5. Partograph Use to Monitor Labor andDelivery (Yapacani Hospital)

Sometimes, however, the team members felt that theimprovement was unsatisfactory. In these cases, the teamswould re-analyze the problem, modify interventions ordevelop new ones, and then re-evaluate the results.

For example, the team at Yapacani studied the results ofthe management action to improve compliance with theuse of the partograph. Although partograph use increasedfrom 0 percent to 69 percent in five months (Figure 5), theteam felt that this improvement was not sufficient, as theresults indicated that 31 percent of deliveries were still notmonitored with the partograph. The team believed thatfurther management action would not yield more improve-

100%

80%

60%

40%

20%

0%0%

42%

64%

Aug,

2000

Perc

enta

ge o

f Del

iver

ies

Corr

ectly

Mon

itore

d w

ith th

e Pa

rtogr

aph

Oct,

2000

Nov,

2000

Dec,

2000

Jan,

2001

Feb,

2001

Sept,

2001

95% 95%

55%69%

Point when QI team

problem-solving

interventions began


ment and therefore returned to analyze the problem as ateam. The team brainstormed potential causes of theproblems and agreed to focus on two: (a) a lack of confi-dence among personnel in their ability to fill out thepartograph correctly, and (b) the fact that the hospital oftenran out of partograph worksheets. The team proposedsolutions to address these causes: training healthcareproviders on how to use the partograph, providing orienta-tion on its importance, and requesting more partographworksheets from the mayor. These solutions increasedcompliance from 69 percent to 95 percent in one month(Figure 5). This example illustrates that QI is a continualprocess where team members continuously seek new waysto improve the quality of care by analyzing for causes andmodifying interventions or developing new ones.

Results

Monitoring indicators to track standards over severalmonths demonstrated the sustained impact of the manysolutions. This section highlights achievements gained bythe QI teams and management action over a four- to six-month period (August or October 2000 to February 2001).

Use of the partograph to monitor labor and deliv-ery. At first, many healthcare providers did not use thepartograph because they lacked confidence in their abilityto use it, did not understand its importance as an earlywarning system to detect complications, and/or could not

use it because worksheets werelacking. The teams implementedinterventions that communicatedto healthcare providers theimportance of the partograph,improved their ability to use it, andensured an adequate supply ofworksheets. The teams measuredhow often the partograph wasused correctly. Correct useincreased from 0 to 95 percent atYapacani and from 24 to 92percent at Buena Vista; SanCarlos achieved 100 percent


Yapacani Buena Vista San Carlos

21

compliance with this standard in two consecutive months(see Figure 6).

This increase in the correct use of the partograph marks asignificant improvement in the quality of EOC—becausecomplications can be detected and referred earlier if thistool is used effectively.

Laboratory exams. All three of the labor and deliveryteams decided to improve compliance with the standard forperforming laboratory exams during labor and delivery, andeach approached it differently. While the San Carlos teamcommunicated with personnel about the importance of thisstandard and conducted a half-day training, the teams atYapacani and Buena Vista established mechanisms thatwould ensure that the tests were conducted even when thelaboratory technician was off duty.

Figure 7. Improvements in Compliance withLaboratory Testing

100%

80%

60%

40%

20%

0%Aug,

2000

Perc

enta

ge o

f Del

iver

ies

Whe

reAl

l Lab

Tes

t Wer

e Pr

ovid

ed

Oct,

2000

Nov,

2000

Dec,

2000

Jan,

2001

Feb,

2001

Sept,

2001

Figure 6. Improvements in the Correct Use ofthe Partograph


100%

80%

60%

40%

20%

0%Aug,

2000

Perc

enta

ge o

f De

liver

ies

with

the

Part

ogra

ph C

orre

ctly

Com

plet

ed

Oct,

2000

Nov,

2000

Dec,

2000

Jan,

2001

Feb,

2001

Sept,

2001


Figure 7 shows their success: increases from 33 to 88percent at Yapacani, from 83 to 100 percent at Buena Vista,and from 7 to 80 percent at San Carlos. These results notonly represent great improvement in providing the appro-priate laboratory tests for women in labor and delivery, butalso demonstrate the powerful effects of learning from eachother through multidisciplinary teams. Here, the laboratorytechnician and nurses contributed significantly to enrich-ing and improving each other’s services provided duringlabor and delivery.

Monitoring during the first two hours afterMonitoring during the first two hours afterMonitoring during the first two hours afterMonitoring during the first two hours afterMonitoring during the first two hours afterdeliverdeliverdeliverdeliverdeliveryyyyy..... All hospitals achieved remarkable improve-ments in compliance with the standard to ensure thatwomen’s vital signs, bleeding, and uterine status areclosely monitored for at least the first two hours afterdelivery. Buena Vista and Yapacani achieved 100 percentcompliance with this standard for two months in a rowwhile San Carlos reached 92 percent (Figure 8). Thesesuccesses in improving immediate postpartum care arenow being recognized throughout Ichilo. In fact, a formdeveloped by the Yapacani team to systematize and recordthis monitoring is now being implemented at all Ichilohealth facilities. This example shows how quality improve-ment can empower and motivate healthcare providers tocontribute to the improvement of the care provided by theentire health system.

Figure 8. Improvements in Monitoring for VitalSigns, Bleeding, and Uterine Status in ImmediatePostpartum Care

100%

80%

60%

40%

20%

0%Aug,

2000

Perc

enta

ge o

f Del

iver

ies

Whe

reAl

l Lab

Tes

t Wer

e Pr

ovid

ed

Oct,

2000

Nov,

2000

Dec,

2000

Jan,

2001

Feb,

2001

Sept,

2001



Protocol of care. A protocol of care is an importantrecord of the treatment provided to a patient. It protects therights of patients by serving as a legal document should areview or audit of a case be necessary to determinewhether proper care was provided. Each hospital devel-oped a format for the protocol of care, clearly indicating thesteps that were taken to care for the woman, by whom, andany necessary follow-up procedures. Figure 9 shows thatYapacani and Buena Vista both improved their perfor-mance in this standard: from 8 to 100 percent and 83 to100 percent, respectively. The San Carlos team suffered asignificant decline in December, indicating a need to revisitthe problem and begin rapid team problem solving again.

Figure 9. Improvements in Completing theFigure 9. Improvements in Completing theFigure 9. Improvements in Completing theFigure 9. Improvements in Completing theFigure 9. Improvements in Completing thePrPrPrPrProtocol of Carotocol of Carotocol of Carotocol of Carotocol of Care Fore Fore Fore Fore Formmmmm

100%

80%

60%

40%

20%

0%Aug,

2000

Perc

enta

ge W

omen

in L

abor

and

Deliv

ery

for W

hom

a P

roto

col o

fCa

re W

as C

ompl

eted

Oct,

2000

Nov,

2000

Dec,

2000

Jan,

2001

Feb,

2001

Sept,

2001


Results related to other labor and delivery stan-dards. In addition to the achievements with the fourstandards addressed at all hospitals and described above,the QI teams achieved results in standards they eachdeveloped for their own hospitals. These standardscovered a variety of areas: infection control, perinatalrecord keeping, and referrals and counter referrals.

The YYYYYapacani and Buena Vapacani and Buena Vapacani and Buena Vapacani and Buena Vapacani and Buena Vista teamsista teamsista teamsista teamsista teams both chose toimplement the standard that all women who deliver at thehospital will have an accurate and complete perinatalclinical record of their pregnancy, delivery, and postpartum


periods.g This record identifies potential risk factors forobstetric and neonatal complications and therefore isessential in planning safe deliveries. Completion of thisrecord has improved at Yapacani from 15 percent (August2000) to 98 percent (February 2001); Buena Vista improvedits performance from 24 percent (October 2000) to 92percent (February 2001).

The YYYYYapacani teamapacani teamapacani teamapacani teamapacani team also improved compliance with threeother standards:

❏ Any woman at risk of an obstetric complication whodelivers at Yapacani hospital will receive a Branulacatheter.8 The QI team studied the use of Branulas andrealized that waiting to insert a Branula until a complicationoccurred meant that precious time was wasted. Further-more, inserting a Branula is difficult during the onset of thecomplication. When the team implemented this standard inAugust 2000, none of the women at risk of a complicationreceived a Branula. The causes for low usage included alack awareness of its benefits, failure to train personnel inusing it, and a shortage of Branula supplies. The QI teamconducted a training and orientation about Branula useand implemented an inventory system to maintain itssupply. Yapacani achieved 100 percent compliance withthis standard from November 2000 until this writing (July2001). This achievement not only indicates effective use ofthe Branula catheter, but also demonstrates the use of theperinatal clinical record and protocol of care to plan forpatients with identified risk factors.

❏ All women referred to a higher level of care during labor,delivery, or the immediate postpartum period will bedocumented in a registry of referrals and counter-referrals.The hospital director sent a memorandum to all healthcareproviders and held a meeting to discuss the importance ofthis process to personnel. At the baseline measurement in

8 A Branula is a flexible, intravenous catheter that can be securely fastenedso that a patient can move her arm freely without having the needle fall outor become obstructed. Branulas were originally inserted for Yapacanideliveries only upon the onset of complications.


August 2000, none of the referrals or counter-referrals atYapacani was registered. Between December 2000 andthis writing, 100 percent of women delivering at the hospitalwere registered; these results are expected to continue.

❏ All women admitted for delivery will receive orientationabout respiratory exercises, positioning, and pushingduring delivery. The team decided to develop a guide forpatients and support materials for healthcare providers toorient women about these topics before delivery. Thelaboratory technician conducts a monthly, random cus-tomer satisfaction survey to assess how many women indelivery receive this orientation and their level of satisfac-tion with it. When Yapacani wrote this standard in August of2000, no women who delivered there had receivedinformation on these topics; since January 2001, 100percent have received this information.

The Buena Vista teamBuena Vista teamBuena Vista teamBuena Vista teamBuena Vista team developed and implemented twostandards to personalize labor and delivery for clients: (a)Each woman is allowed to choose her position duringdelivery, and (b) Each woman is presented with the optionof being accompanied by a family member during delivery.

To implement these standards, the team sought training forhealthcare providers in helping women to use alternativepositions for delivery and purchased a special supportcushion for squatting deliveries. The team also boughtfabric to make extra gowns so that a family member couldattend normal deliveries. As of February 2001, 100 percentof women in prenatal care at Buena Vista were presentedwith the option of choosing an alternative position (46percent elected this option). Likewise, 100 percent of thewomen were presented with the option of being accompa-nied by a family member during delivery (53 percent chosethis option). A QI team physician indicated that for her, thiswas one of the most powerful changes implemented by theQI team. The physician told of a woman who delivered herfirst six children at home with a traditional birth attendant.The physician said that the woman chose Buena Vista forher seventh delivery because she could choose herposition and have her husband with her during delivery.


This story illustrates one of the main principles of qualityimprovement: to provide quality care with a focus on theclient.

The San Carlos teamSan Carlos teamSan Carlos teamSan Carlos teamSan Carlos team concentrated on compliance with astandard to address infection control: “All women in laborwill be prepared for delivery with warm water and soap.”When the QI team measured for the baseline, only 4percent of the women in labor received this care. The teamdecided that since this was a relatively simple procedure,communication from the hospital director to personnelwould improve compliance. Indeed, in January andFebruary 2001, San Carlos achieved 100 percent compli-ance.

Next Steps and New Challenges

The labor and delivery QI teams accomplished rapidresults in improving compliance with standards. All of theQI teams have decided to continue to monitor the stan-dards discussed in this case study at least quarterly toensure that the results are sustained. In addition to sustain-ing improvements, the QI teams are sharing successfulinterventions with each other and adapting interventionsthat would improve the quality of their services.

The next challenge will be to maintain these results whilecontinuing QI efforts in new standards and technical areas.The teams plan to focus on improving compliance withadditional EOC standards, particularly in the managementof obstetric complications.

Quality Improvement Insights

The improvements highlighted in this case study—in-creased compliance with standards of care for labor anddelivery—will contribute to the overarching goal of theteams: to reduce maternal mortality and morbidity in Ichiloby providing quality care in their hospitals.

The teams in Ichilo have demonstrated their ability to applythe QI methodology to improve compliance with standardsfor labor and delivery. Once decisions were made on whichproblems could be addressed by management action and


which required team problem solving, an ambitious numberof areas for improvement could be addressed. By closelymonitoring each standard, team members determined theimpact of the interventions and used the data to determinewhether further interventions were necessary. The Ichiloexperience demonstrates how multidisciplinary teams canuse rapid team problem solving to achieve dramaticimprovements in the quality of care. It also provides severalkey insights about future applications of this methodology:

Improving the quality of care does not alwaysrequire additional resources. Dramatic improvementswere achieved by changing the way work was organizedand without adding new resources. An example is inlaboratory testing for patients in labor and delivery whenthe technician is off-duty. Instead of giving up becauseanother laboratory technician could not be hired, the teamprobed for creative, low-cost solutions and realized thattraining nurses in laboratory tests was a viable alternative.

Quality improvement fosters a sense of owner-ship of the problem and solutions. Many solutionsare continuing because the healthcare workers feelownership of the interventions. This sense of ownershipstems from the fact that the team members worked togetherto identify and analyze a problem that they determined wascritical to the quality of care and then developed, tested,and implemented the interventions.

Quality improvement is motivating, especiallywhen it is noticed and rewarded by clients,managers, and others. In April of 2001, the MOHranked Yapacani as the fourth best hospital in Bolivia (outof 2,500) for secondary-level care. Based on its high levelof compliance with standards and customer satisfaction,this recognition validated the team’s efforts and motivatedthem to continue to expand quality improvement endeav-ors.

Quality improvement is most successful whenteams have the full support of leaders. Hospitaldirectors, district directors, and SEDES supervisors wereactive in the QI process; some even served as team


members or facilitators. By showing interest in the teams’work and communicating that quality improvement was apriority in Ichilo, these managers helped sustain team effortand contributed to their success (and continue to do so).

Quality improvement teams can achieve rapidresults. When team members have insight into thepotential causes of a problem—or determine that manage-ment action is sufficient—problems can be analyzed andsolutions developed and tested quickly. The QI teams inIchilo exemplify this point as they achieved results as earlyas two months after implementing solutions. These rapidresults motivate QI teams as they can quickly see thebenefits of their hard work.

Sharing ideas and results among teams leads torapid results. The teams in Ichilo achieved rapid resultsin part because they focused on many of the same issuesand could share interventions with each other and adaptthem for their own hospital’s situation. In February of 2001,the QA Project coordinated a workshop for teams to presenttheir results and share their interventions. Teams learnedfrom each other’s experiences and developed a healthyspirit of competition to demonstrate the greatest improve-ments in quality. For example, the efforts to personalizelabor and delivery at Buena Vista sparked interest amongthe other teams in implementing similar standards.

Decentralized health systems seem to present aresponsive environment for the application of theQI methodology. The highly decentralized system inBolivia fostered active participation among personnel anddistrict managers. Since district and departmental supervi-sors are responsible for the management of local healthservices, they had the authority to lead and support thequality improvement teams.

Facilitation was critical at two points in theBolivia experience: at the beginning of rapidteam problem solving and at decision-makingjunctures. The facilitators helped determine whethermanagement action or team problem solving would be thebetter approach for different problems. They urged teams to


adopt a team spirit rather than blame others when thebaseline measurements were disappointing. Facilitatorsalso participated in discussions evaluating the feasibility ofproposed solutions and helped monitor indicators.

Rapid team problem solving can bring communityconcerns to healthcare quality improvement. Theteams both addressed MOH standards and developed newones that reflect their understanding of clients’ expecta-tions. Satisfying clients fosters healthcare utilization and,ultimately, better health outcomes.

a World Bank. 2000. Entering the 21st Century: World Development Report1999/2000. New York: Oxford University Press.

b Pan American Health Organization. 1999. Situacion de Salud en Bolivia.<www.ops.org.bo>.

c J. Ashton. 2001. Monitoring the Quality of Hospital Care. Health Manager’sGuide. Bethesda, MD: Published for the U.S. Agency for InternationalDevelopment (USAID) by the Quality Assurance Project.

d Massoud, R., K. Askov, J. Reinke, L. Miller Franco, T. Bornstein, E. Knebel,and C. MacAulay. 2001. A Modern Paradigm for Improving HealthcareQuality. QA Monograph Series 1(1). Bethesda, MD: Published for the U.S.Agency for International Development (USAID) by the Quality AssuranceProject.

e Hermida. 2001. QAP Results from the Field: A Quarterly Report on Resultsfrom Latin America. Quito, Ecuador: Published for the U.S. Agency forInternational Development (USAID) by the Quality Assurance Project.

f W. Shewhart. 1931. The Economic Control of Quality of ManufacturedProducts. D. Van Nostrand Co, New York: Reprinted by the AmericanSociety of Quality Control in 1980.

g Díaz, A.G. 1990. Sistema Informático Perinatal. Centro Latinoamericano dePerinatología. Montevideo: Uruguay.


Improving Compliance with Standardsfor Essential Obstetric Care in Bolivia

This case study highlights the effectiveness—and describes the process—of rapid teamproblem solving in a developing countryhealthcare setting. It tells the story of qualityimprovement teams comprised of labor anddelivery personnel at three hospitals in Ichilo,Bolivia. The teams applied rapid team problemsolving in areas related to essential obstetriccare and achieved significant results in only afew months. With support from the Santa CruzHealth Department and the Quality AssuranceProject, these teams identified opportunities forimprovement related to standards that had beenpromulgated by the Ministry of Health or thatthey devised themselves to reflect the qualityexpectations of their communities. Particularlyimpressive results were achieved with regard tousing the partograph correctly during labor anddelivery and in postpartum monitoring, both life-saving protocols.

This study details the steps of rapid teamproblem solving as performed by teams,presents the indicators of improvementsachieved, and offers insights for other groupsseeking to implement quality improvements byusing rapid team problem solving at their ownhealthcare facilities.


Documents

Improving Compliance with Standards for Essential ...pdf.usaid.gov/pdf_docs/PNACQ767.pdf · cases, more than one QA ... Improving Compliance with Standards for Essential Obstetric