||Dennis Chao1, A.K. Nandakumar2, Abul Barkat3, Avijit Poddar4
||1RTI International, North Carolina, United States 2Brandeis University, Massachusetts, United States. 3Dhaka University, Dhaka, Bangladesh. 4The Human Development Research Centre, Dhaka, Bangladesh
|1st country of focus:
|Relevant to the conference theme:
||New roles and responsibilities of health personnel
|Summary (max 100 words):
||Short patient contact time (PCT), or the time health providers spend with patients, adversely affects the quality of health services in most developing countries. Building on previous studies of health care quality, this study investigated determinants of PCT using data from a nationally representative sample of 130 nongovernmental organization clinics in Bangladesh that provided essential health services to poor and nonpoor patients. PCT was examined for providers with similar competencies and working conditions. Regression analysis results identified several factors that significantly influenced PCT: number of patients, providers’ direct services time, type of service, and whether patients paid for care.
|Background (max 200 words):
||Quality of health care services is an issue of widespread concern in both developed and developing countries. Medical practitioners and researchers alike are seeking answers to questions such as how to measure such quality, what factors affect quality of care, and what types of changes can leverage the greatest improvements in quality. The quality-of-care issue has particular urgency in countries with large populations living in poverty: Even if health services are free, the poor and marginalized are unlikely to seek them if the quality is poor. While many factors affect the quality of services, an emerging body of literature suggests that the amount of time providers spend with patients (known as patient contact time, or PCT) is a good proxy for quality and a basis for patients’ satisfaction. The reasoning behind this is that short PCT is often related to misdiagnosis and poor management of cases. Similarly, low PCT leads to patient dissatisfaction. Given these findings, a logical argument is that optimizing PCT not only will improve quality of services but also will lead to enhanced patient satisfaction. Understanding the factors that affect PCT, especially with regard to services provided to the poor, is of critical importance.
|Objectives (max 100 words):
||The main objective of this study is to further investigate factors that determine the quality of health service, with data from a nationally representative sample of non-governmental organization (NGO) clinics in Bangladesh that provided essential health services to poor and non-poor patients. We followed the approach taken in a Paraguay study, but extended the scope in several directions. First, we focused on services provided by providers with similar competence in facilities managed by organizations with similar structures and systems. When all providers possess the same competence, it is reasonable to assume that the length of provider-patient interaction time itself may be a good approximation of the practice factor. This allowed us to concentrate on the analysis of patient contact time (PCT) as a proxy for the quality of health service and to identify factors determining PCT that could lead to program and policy interventions to improve PCT and quality of service. Second, using PCT as our indicator also allowed us to bring allocation of providers’ time into the analysis and to view patient contact time in a framework of providers allocating total available time in the clinics and patients competing for their direct service. Third, the Bangladesh dataset contained information on patients’ payments for services and therefore provided an opportunity to examine the discrimination factor through patients’ ability to pay. We also intended to use the Delphi method to develop a “standard contact time” (SCT) for the services studied. Standard contact time is the length of time that experts believe health providers will need if they follow protocol and carry out appropriate steps with patients. Providers’ knowledge about and interpretation of SCT is one factor that influences PCT in a clinic setting. With the matched information that we had on both the provider side and the patient side, we expected to gain new knowledge regarding the PCT of essential health services in a developing country and factors that influence the actual PCT. Our findings should inform policy and program interventions aimed at lengthening PCT to improve quality of care.
|Methodology (max 400 words):
||Model Providers have a finite number of working hours in a day to allocate among, administrative work, rest breaks, and direct service provision. We hypothesize that four factors will impact their allocation choices, particularly for PCT. The first factor is the time available for providing direct services. We hypothesize that time available for direct services is positively related to the time providers actually do spend with patients. The second factor is the number of patients seen during a day. We hypothesize that there is a negative relationship between the number of patients seen and PCT. The third factor affecting PCT is the complexity of the service and the optimal time providers should be spending on these services. Standard contact time reflects the number of procedures recommended by established protocols as well as the time needed to undertake these procedures. We posit that everything else being the same, longer SCT will be positively related with PCT. The fourth and final factor is whether the patient pays for services. We posit that there is a positive relationship between PCT and the payment status of the patient and that PCT for paying patients will be longer than those of nonpaying patients. Data We tested our hypotheses with data from a cost structure and staff utilization survey conducted for the Bangladesh NGO Service Delivery Program (NSDP) in 2005. A central aspect of the study was a measurement of the amount of time that health providers spent in consulting and treating patients. In our study, PCT was defined as the time spent in direct contact with patients for (1) taking histories and reviewing records; (2) carrying out physical examinations; (3) preparing for and conducting procedures; (4) writing notes and prescriptions; (5) making referrals; and (6) advising, instructing, and/or counseling patients. The time-and-motion approach was used in this study to measure PCT. NSDP delivered primary health care services through an umbrella of NGOs. The network included more than 30 NGOs and operated 318 static clinics and 8,200 satellite clinics throughout Bangladesh. Sample In total, 138 clinics were randomly selected to reflect the universe of NGO clinics supported under NSDP. Of the 138 clinics studied, 60 were urban static clinics, 26 were rural static clinics, 30 were urban satellite clinics, and 22 were rural satellite clinics.
|Results (max 400 words):
||Descriptive Analysis: The mean PCT was 8.07 minutes. On average, providers spent nearly 5 minutes, or 37%, less time than SCT. Despite the below-SCT averages that the study revealed, Bangladesh ranks highly among low-income countries in PCT. We estimated the distribution of two major provider characteristics: (1) the total time available for serving patients and (2) the number of patients served in a day. We observed a wide range of values for these two variables and found that they were highly correlated to PCT. The results also provided information on the economic status of patients and whether they paid for services received. From the total sample, 69% of patients reported having paid for care. The average amount paid for care was 9.18 taka. The correlations between PCT and payment were significant. Regression Results: A regression analysis was conducted to test whether providers adjusted their practice patterns based on the socioeconomic status of the patients and whether patients paid for services. Similarly, we tested the hypothesis that the number of patients seen and the available time for direct services affected the time provided for each patient. We found the regression coefficients for both the patients served and available time variables to be of the right sign and highly significant. A 1% increase in patients served would reduce PCT by 0.34%. On the other hand, a 1% increase in available time for direct service would increase PCT by 0.22%. The regression results also suggested that the longer the SCT is the longer the PCT is likely to be. On the patient side, an important finding was that PCT responded to payment status, which suggests that providers did adjust their behaviors based on whether patients paid for services. Simulation: We used the regression results to simulate the impact of changes in other variables on PCT. The regression equation predicted PCT at 8.07 minutes. We calculated the ideal time available for clinical care by assuming that all providers would actually show up for work and stay the entire workday. We also assumed that all patients would pay the average of actual user fees paid in the sample. Under this scenario, the projected PCT was 9.20, which was 14% higher than the original PCT. The simulation results clearly demonstrated that it would be possible to increase PCT by a combination of supply and demand policy changes.
|Conclusion (max 400 words):
||This study is the first time a systematic analysis has been conducted of factors in Bangladesh affecting PCT. By employing a large and random sample and using the most reliable of the three suitable survey methods, this study obtained estimates of PCT of essential health services supplied by providers of similar competence and compensation. The analysis of the means and variances of PCT significantly extended our knowledge of the importance of PCT and our understanding of the factors that influence it. Results of the regression analysis yielded new evidence on the factors from both the supply and demand sides that contributed to the variations of PCT among health providers with comparable competence. On the supply side, we found that, as expected, both the time available for direct services and the number of patients influenced PCT. On the demand side, we found whether or not patients paid for care was significant in determining PCT: paying patients received more time than nonpaying patients. The evidence also supports the principle of a “payment follows service” approach in strengthening the position of the poor in receiving free services. The results from this study and the simulation exercise also have policy implications. There are policy and program interventions that can effectively increase the PCT for essential health services to approach the SCT. For example, on the supply side, it is important to reinforce the regulations that require providers to spend their full working hours in the clinics. In addition, training and communication about standards of care, especially SCT, could rise providers’ awareness of both what is expected in terms of best practices and what effects PCT has on health outcomes. On the demand side, an alternative financing approach may be required to further improve the quality of services to the poor. Our results suggest that increases in PCT, and therefore quality of care, can be achieved by paying physicians for providing free care to the poor as opposed to requiring providers to waive user fees for the poor. In other words, instead of using a supply-side grant to provide free services to the poor, it would be better to use demand-side financing to enable the poor to pay for services. With these changes, it should be possible for PCT at NGO clinics in Bangladesh to reach levels observed in developed countries such as the United Kingdom.