Background

The 2016 outbreak of Zika virus across twenty countries in the Americas and the preceding Ebola outbreak of 2014–15 in West Africa have underlined the importance of routine disease surveillance in an increasingly interconnected world [1]. Both these epidemics also exposed the inability of the fragile public health systems within the countries to respond swiftly or to preempt the scale of the problem [2, 3].

Government spending on health is alarmingly low in a majority of low and middle-income countries, which leaves the public healthcare system chronically underfunded and impoverished [4]. Often forcing them to make compromises on the quality of care and fueling issues with patient satisfaction and acceptance [5]. Furthermore routine surge functioning over years leaves them vulnerable to breakdowns at the lowest level of imbalance. In the backdrop of these infrastructural and financial challenges within the public healthcare sector, the private sector has gained strength globally [6]. Private health service provision is significant and dominant particularly in urban areas [7, 8]. Currently, more than half of the global population lives in urban areas. This number is expected to rise to 66%9 as the population races to 9.7 billion by 2050 [9, 10]. We will probably witness a simultaneous and proportionate surge in both the demand and supply of private healthcare in both rural as well as urban areas globally.

Private practitioners are the preferred first point of care in emerging economies because of perceived quality, lower costs, speedy care, flexibility of payments, and accessibility [5, 11]. They already account for over 50–80% of the out and in-patient care in countries like India, China, South Africa, Brazil, and Nigeria amongst others [7, 12, 13] In their position private primary care practitioners are best placed for the early detection of outbreaks as well as routine monitoring of disease trends. However, their role in current disease control programs is largely limited to service provision and outbreak response [14]. Their contribution to health information systems is in most countries marginal and largely voluntary, leading to gross misrepresentation and underestimation of disease burdens [15, 16]. Therefore involving the private sector in routine disease surveillance is no longer a choice but a necessity.

While public-private partnerships can be cost-effective [17] and have been widely and successfully implemented in several individual disease control programs e.g. HIV/tuberculosis, polio, malaria etc. [18], their potential for routine disease surveillance remains largely untapped. The objective of this systematic literature review is to explore the experiences made across emerging economies with regard to the role, contribution, and involvement of private practitioners in disease notification. We examine the factors that determine the inclusion of and the participation thereof of private practitioners in surveillance activities.

Methods

A systematic literature search was conducted in September and October 2015 with an additional update in February 2016 using the databases PubMed and Web of Science. Grey literature searches were conducted using WHOLIS (Library and Information Networks for Knowledge Database) WHO-IRIS (Institutional Repository for Information Sharing) and the CDC Stacks databases. The PRISMA Statement checklist for systematic reviews was referred for the review process [19].

Inclusion criteria were set at peer-reviewed, and grey full text empirical, original articles in English with no limits for year of publication or study design. The key search terms used in permutations and combinations included “private practitioner” (“private practice”, “private sector”, “private healthcare provider*”, “private facility*”, “non-public sector”, “non-public physician”, “for profit sector”, “for profit facility” “private physician”) and “disease surveillance” (“public health surveillance”, “sentinel surveillance”, “population surveillance”, “epidemiological monitoring”, “information systems”, “hospital information systems”, “health information systems”, “management information systems”, “ambulatory care information systems”, “automatic data processing”, “electronic health records”). Search algorithms always included terms related to private practice and disease surveillance (Additional file 1). Search results were merged using EndNote X7 and duplicates removed.

All articles (136) selected on the basis of title and abstracts were retrieved. Manual screening of reference lists identified eight further articles. A total of 144 articles were reviewed full text and data extracted by two reviewers (CB and KB) and verified independently (RP). Disagreements (13 articles) were resolved with mutual consent. Studies that explicitly investigated the role of private practitioners in routine disease surveillance were conducted in low and middle-income countries as defined by The World Bank [20] were included in the final review. The main reasons for exclusion of 94 articles (Fig. 1) included a lack of focus on routine disease surveillance, studies from high-income countries followed by non/full text opinion or review papers.

Fig. 1
figure 1

PRISMA flow diagram summarizing the literature search process

Full size image

Limitations of the review

The searches were conducted in four databases and limited to English language only which may have led to a degree of selection bias. We also used a narrow definition of private sector to include only practitioners excluding NGOs, laboratories, and pharmacies and the informal sector (unregistered or untrained) stakeholders, which limits the generalizability our findings to the private sector. We limited our search to emerging economies. Important lessons could have been drawn from experiences in the developed economies. About a third of the studies were intervention studies, which could have influenced the reporting behaviors of the practitioners. Additionally the studies were conducted in very different settings, at different scales for single and multiple diseases with both mandatory and voluntary reporting systems for different periods of observation as well as differing study designs. The heterogeneity of the studies meant that we were unable to perform in-depth analyses of the factors and could not draw generalizable inferences. For the purpose of the review we extracted data on the lessons learned and factors reported to influence reporting as identified in these studies and aim to present a summary of the facilitators and barriers to private sector involvement in routine surveillance. There were no major differences across the findings from these studies and the observations are of value in shaping the discussions and identifying specific areas of in-depth research in the future.

Results

Overview of the studies

Forty studies were included in the review (Table 1). In addition, we found ten review papers that provided detailed recommendations. We present them as supporting studies only (Table 2). The review summarizes the experiences drawn in 28 countries. The studies span 13 countries in Asia (Bangladesh, China, Iran, India, Indonesia, Malaysia, Myanmar, Nepal, Pakistan, Philippines, Taiwan, Thailand, and Vietnam), six in Africa (Ethiopia, Kenya, Morocco, Nigeria, South Africa, and Uganda) and nine in the South American region (Bolivia, Costa Rica, Dominican Republic, El Salvador, Guatemala, Honduras, Mexico, Nicaragua, and Peru).

Table 1 Overview of the studies and their main findings
Full size table
Table 2 Overview of the supporting studies
Full size table

All the studies except one [21] were single country studies conducted between 1996 and 2015 and a majority of them were from India (15, 37.5%) [14, 22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37]. The scale of the studies varied from small city areas to the national levels, but about half (19, 47.5%) of them were conducted at an individual city level [11, 16, 17, 22, 23, 28, 31, 33, 37,38,39,40,41,42,43,44,45,46].

Fifteen (37.5%) studies were mixed methods surveys. Eleven (27.5%) were intervention studies targeted at improving private practitioners participation in disease control programs [11, 17, 21, 22, 28, 29, 40, 41, 47,48,49]. Nine studies (22.5%) used secondary record review to evaluate disease notification, and four (10%) of them were Knowledge Attitude and Practice (KAP) surveys. The average private practitioner response rate (in 20 studies) was 78% and ranged from 1.5% in Nigeria [50] to 100% in Iran [51], Bangladesh [46], and Uganda [52]. Three studies from India [29, 30, 37] reported a higher response rate to interviews compared to requests of record reviews.

All except one study [34] referred to communicable diseases (STIs, malaria, typhoid, influenza and five studies [26, 33, 51,52,53] with multiple diseases) and a majority (28, 70%) were related to tuberculosis. Eighteen studies (45%) refereed to laboratory confirmed case detection and ten (25%) referred to both suspected as well as laboratory confirmed cases. Only one addressed syndromic surveillance for influenza [54]. Eight studies (20%) [15, 16, 27, 29,30,31, 52, 54] investigated private practitioners reporting behaviors towards voluntary case reporting. Thirteen (32.5%) [17, 24, 28, 34, 37, 38, 40, 41, 44, 50, 53, 55, 56] investigated systems that expected mandatory case reporting from the private sector.

Fifteen studies (15, 37.5%) [16, 22, 25, 27, 28, 34, 38, 40, 43, 45, 46, 48, 55, 57, 58] stated that involving private sector reduced diagnostic delays and improved case detection (7 to 50%). Even when only a fraction of private practitioners became active, the case detection rose significantly [22]. Involving the private practitioners in surveillance activities also helped identify an emerging disease (leptospirosis) in India [26], recognize patterns in health seeking behaviors in China [42], Nigeria [50], and Morocco [54] and detect comorbidities in Kenya [38].

Barriers to notification

Eight (20%) [15, 16, 30, 33, 37, 49, 51, 53] studies identified knowledge of the practitioner as the most important determinant of case detection, confirmation and notification (Table 3A). Knowledge about disease control programs and their diagnostic requirements was higher amongst public as compared to private practitioners [36]. Although the knowledge regarding disease detection was lower amongst alternate medicine practitioners as compared to allopath practitioners in India, the surveillance practices did not differ significantly [14]. General practitioners were more likely to be aware about the importance of notification than specialists in India [29, 30] and Malaysia [15]. The duration (>5 years) of practice (OR 11.4, 95% CI 1.99, 65.58, p = 0.001) was significantly associated with practitioner reporting in Uganda [52]. In Malaysia, practitioners were willing to report cases only when they were confident of their diagnosis [15].

Table 3 Barriers to case reporting at the practitioner and government/public sector end as identified by the studies
Full size table

Apart from disease knowledge a clear understanding of the notification procedures is critical to reporting. Lack of clear instructions, inadequate dissemination of guidelines and no assistance with reporting procedures, supervision or feedback were identified as the most important reasons for under reporting in nine studies (22.5%) [14, 22, 24, 33, 34, 43, 46, 48, 51] (Table 3B). Over 50% of the 238 practitioners in the study from Malaysia did not know whom to report to and did not have reporting forms [15]. In Uganda about half (49%) of the 109 practitioners knew where to send the report but only 21% (n = 23) knew which form to use [52]. Practitioners who considered the system inconvenient or were unfamiliar with the reporting procedures were less likely to report cases in Taiwan [53].

Simplicity of reporting procedures and the mode of reporting offered (telephone, email, paper etc.) were important determinants in India [33] and Taiwan [53]. Findings from Bangladesh [46] and India [37] indicate that reporting compliance increases over time as the familiarity to the system grows and therefore sustained efforts are necessary. The rated public health importance of the disease (e.g. cholera > enteric fever) or sense of emergency (e.g. Zika) also affects disease notification from the private practitioners [26].

Appreciating the importance of surveillance or adequate knowledge of the disease and its reporting procedures does not ensure case notification. Two studies in India suggested that only about half of the practitioners who understood the importance of notification agreed to report/participate in surveillance activities [30, 33]. Yeole et al. (2015) support these findings [37] where although 64% of 831 practitioners agreed to participate in a surveillance system, only 16% (n = 87) actually notified data. Although about half of the participants (104, 46%) in a study in Mumbai said that there were no barriers to reporting, this was not reflected in their reporting practices [33].

The attitude of the practitioners was reported as significant determinant of case notification in studies from Malaysia [15], India [36], and Uganda [52].

Of the 238 practitioners in a study in Malaysia, 73.1% exhibited unsatisfactory “attitude” towards disease notification, 81.9% showed complacency, 66.9% ignorance and 23.5% indifference respectively [15]. Ullah et al. (2012) report it was the negative attitudes of the practitioners towards the government officials rather than the notification process itself that affected reporting in Bangladesh [46]. Ahmadi et al. (2013) note that it was the negative attitude of the data collectors towards data compilation and towards the practitioners, which discouraged practitioners from notification in Iran [51].

Lal et al. (2011) and Ambe et al. (2005) conclude that there is mutual distrust and prejudices and suggest that trust building is necessary at both the public and private practitioners end alike. Seven (17.5%) studies [14, 22, 30, 37, 46, 48, 51] and six (15%) studies [28, 31, 34, 48, 55, 59] identified the lack of coordination/collaboration between the government and the private sector as the main barrier for case reporting respectively (Table 3B). There is a need for managing perceptional conflicts at both ends [30]. While the government sector should attract and sustain private practitioner attention, the private practitioners should exhibit their responsibility towards disease notification [22, 28]. Five (12.5) studies [11, 14, 15, 51, 59] suggest that the motivation of the practitioners also played an important role in disease notification.

While eight studies (20%) [14, 15, 30, 34, 37, 53, 56] and four (10%) studies [14, 15, 33, 51] identified breach of patient confidentiality and legal issues as reasons for not reporting respectively, Philip et al. (2015) suggest this is a perception only as legal frameworks in most countries allow case notification to governments [30].

Nine (22.5%) studies [14, 17, 27, 31, 32, 48, 50, 51, 58] reported the lack of adequate and skilled staff and equipment (e.g. computers) as the main barriers to case reporting. Interestingly only seven (17.5%) studies [14,15,16,17, 33, 34, 53] identified lack of time as the main barrier. While the maintenance of records within individual facility was significant determinant of case notification in Ethiopia [60], the availability of information materials, registers and reporting formats affected case reporting in Uganda [52]. Access to a laboratory (OR 3.79, 95% CI 0.99, 14.55, p = 0.05) played an important role in the willingness of private practitioners to report cases in India [17].

Recommendations to improve reporting

Obtaining an overview of the private sector by identifying the different actors and clearly stating their roles and responsibilities was recommended as the first step towards government regulation of the private sector (Table 4) with regard to disease surveillance [11, 22, 27, 28, 31, 45, 52]. Krishnan et al. (2006) suggest a separation of the governments’ regulatory function from public service provision to ensure strict action against reporting defaulters [27]. Eight studies each recommended standardization of unified reporting procedures [11, 22, 31, 33, 46, 48, 51, 57] and earmarked public financial resources for capacity building within the private sector for disease surveillance [14, 17, 21, 23, 31, 40, 52, 55].

Table 4 Recommendations to improve private practitioner participation in disease surveillance
Full size table

At the district administration level, the main recommendation was to provide surveillance training (regular [14, 15, 21, 25, 28,29,30,31, 33, 34, 36, 46, 48,49,50,51,52, 55, 57, 60] and continuing medical education [14, 29, 30, 36, 37]). Five (12.5%) studies recommended revision of regular undergraduate and postgraduate medical curricula to incorporate a stronger focus on surveillance activities [15, 28, 29, 51, 57]. Eleven (27.5%) studies [16, 29, 31, 34, 36, 37, 40, 50, 52, 57, 60] recommended that IEC materials and guidelines should be widely disseminated. Eleven (27.5%) studies [27, 30, 32,33,34, 37, 43, 45, 46] recommended improving the communication between the government, public and private sectors (Table 4).

At the program level, feedback (10, 25% studies) [11, 14, 16, 17, 26, 27, 31, 45, 46, 52] and supportive supervision (9, 22.5% studies) [25, 27, 33, 45, 46, 48, 49, 51, 52] were the main recommendations. Eight studies (20%) [14, 16, 17, 27, 29, 40, 51, 53] recommended that in the absence of regulation there may be the need to provide an incentive/reward (e.g. CME credits) for reporting. Only one study suggested that the incentives could be non-financial (e.g. technical assistance, supervision, free diagnostics etc.) [14, 18].

Discussion

Findings of the review suggest that the knowledge, attitudes, perceptions, and practices of the government, the public sector practitioners (who implement surveillance programs), and the private sector practitioners affect case notification. Interventions targeting all three groups are therefore mandated. Training, timely feedback, and regular supportive supervision from the program managers can improve case notification. Simple standardized reporting procedures with clear guidelines and effective communication between those providing and receiving the data are key to effective private practitioner participation in disease surveillance. More importantly regular publication of disease data from both public and private sectors can encourage reporting as it serves as feedback to those reporting. Improved representativeness serves as evidence base and aids the use of data for decision making at the national levels.

The factors affecting case reporting from the private practitioners as identified in this review are similar to those reported for under reporting from the public sector within the Integrated Disease Surveillance and Response systems (IDSR) which collect routine data mainly from the public sector [61, 62]. Therefore an important first step would be to make the distinction between the roles of the policy makers, healthcare providers and people (individual practitioners)! There is the urgent need to separate the words public sector and government [27]. While public sector facilities and practitioners funded through general taxes are responsible for implementing disease control programs in most LMICs, the role of the government is wider. Governments bear the responsibility to regulate and develop both the public as well as the private sectors equally [5]. All data collection should be substantiated with follow up action in order to encourage reporting compliance in both the public and private sectors. Furthermore involvement of the end users (patients and communities), non-government and civil advocacy organizations should be considered to create enabling environments for disease data reporting.

Heterogeneity of private sector [22, 45] and the varying size, arrangements and functionality of the informal sectors [40] are probably the main reasons why it is a resource and logistical challenge to regulate the private sector in most LMICs [31]. Often records of all healthcare providers are not available even with the government which is a challenge for assessing reporting consistency and disease burdens [14]. Governments should exercise authority and use registration as a tool to regulate the formal private sector facilities as a start. Although troublesome to implement, a “single registration platform” for all providers should be advocated to record the legal status, size, and nature of services provided [63]. Stronger legislative and regulatory frameworks are necessary to harmonize the diverse set of stakeholders [6, 17].

The next step would be to simplify and standardize the reporting procedures, provide clear guidelines and ensuring that the information reaches every intended private practitioner [18, 64]. Where possible, data should be obtained in a format (email, toll free number, text-messages, paper forms etc.) preferable to the practitioner and without incurring extra work or interrupting workflows [33]. Private sector should be involved in decision-making processes with regards to surveillance and invited to become members of national policy implementation boards [13, 65].

Human resource development in the private sector should be seen as important as in the public sector. Governments should invest in the training of private practitioners (CME, workshops etc.) and accessory staff in surveillance procedures. Medical curricula should be revisited to emphasize the importance of reporting [13]. A substantial “knowledge-application/practice gap” exists and even amongst practitioners who know and understand the importance of reporting only about a half will notify [17]. Therefore intermittent sensitization campaigns highlighting the importance and processes of surveillance are useful [66]. Small individual clinics constitute a major proportion of the private sector and the public sector should have sufficient capacity to train and supervise the large (and growing) numbers [59].

In addition, supportive supervision through onsite visits by program officers facilitate trust building and should be implemented [67]. While this is resource intensive, extensive initial input followed by regular follow up pays off in the long run [37, 46]. “Without feedback, practitioners soon realize that it makes no difference whether they report. Information is of value only when it is used for decision-making” [68]. Regular feedback (telephone, newsletter, review meetings etc.) is an essential determinant of provider motivation and should be mandatorily provided [69,70,71].

Four [35, 72,73,74] studies suggested that reporting should be mandatory and a penalty be imposed for non-reporting. However voluntary systems are more accurate and although legal obligation evokes reporting, reporting is more complete when the practitioner understands the importance of sending reports [75, 76]. Six studies [16, 17, 27, 29, 40, 51] from India, Iran and Pakistan suggested that in the absence of regulatory frameworks financial incentives may become necessary for consistency in voluntary reporting. While there are fundamental differences in the values of individuals in the not for profit public and the only for profit private sectors that are difficult to navigate, this approach is not sustainable [77]. Awarding non-financial incentives such as free training, credit points towards Continuing Medical Education (CME) for consistent reporting, free drugs, laboratory tests, and access to scientific journals or books etc. may be offered. Infrastructure assistance such as software loaded basic handheld devices with a direct reporting tools could be an option for the future [5, 66]. Alternately commissioning surveillance from large private hospitals with closely monitored contracts and performance-based incentives may prove beneficial [6]. Once routine data reporting improves, efforts to investigate the quality of the data received from private sector and comparative assessments between public and private providers could be useful.

The links between the governments, public, and private sectors need to be identified and strengthened. Intermediary bodies such as Non-Governmental Organizations [25, 28, 31], professional associations [25, 64], private labs [35, 78], private medical colleges [13], and pharmacies [73, 74] are crucial connectors and can facilitate communication between the three groups [35, 72,73,74]. Practitioners in LMICs often work in both the public and private sectors. Those with dual roles could be roped in to initiate a dialogue with the private sector and also promote peer reporting from defaulting practitioners [65]. Practitioners’ ranking based on consistency in routine disease data reporting has been attempted within the Integrated infectious Disease Surveillance and Response system (IDSR) in India and could be tested at scale in other settings. Most importantly all actors need to acknowledge the overarching role of the government to regulate and support them at the same time. Sharing of knowledge and best practices alongside skills transfer should be encouraged across regional and national governments in order to avoid delays in implementing changes already tried and successful in better involvement of the private sector in routine disease surveillance.

Conclusion

The current private practitioner participation in disease surveillance efforts is appalling. The main barriers to their participation are inadequate knowledge leading to unsatisfactory attitudes and misperceptions that influence their practices. Complicated reporting mechanisms with unclear guidelines along with unsatisfactory attitudes on behalf of the government also contribute to the under reporting. Infrastructural barriers such as availability of computers and human resources need rectification. Governments need to take leadership and foster collaborative partnerships between public and private sectors for routine disease surveillance and exercise authority when needed. Surveillance program officers need to provide periodic training, offer supportive supervision and regular feedback to practitioners from both public as well as private sectors.