The present study followed a mixed method strategy that contributes to improving knowledge of activities undertaken by AIDS-NGOs and serves as a basis from which to develop a final proposal of indicators to assess these prevention-control based activities. As shown in the results, this methodological strategy implies that the different phases complement one another and progressively richer information is obtained [20, 37].
The results allowed us to know all the preventive-control based activities done by AIDS-NGOs in Catalonia and how they could be assessed. The first two phases showed us a relationship between objectives and activities, and what other activities were not registered in the AIDS-NGOs annual reports. The qualitative phase showed how the preventive-control activities are done and what the main barriers and limitations to evaluate it are. Finally, the indicators consensus meetings enabled us to talk with AIDS-NGOs professionals and establish a consensus on the indicators proposed to evaluate the activities in the daily work flow.
The preventive activities of the AIDS-NGO are based on a participatory health education model adjusted to the people's needs, based on empowerment and focusing on the ideas of knowledge and skill. This education strategy follows the principles put forward by the WHO .
Results emphasise the importance of offering ancillary services  to people in need of HIV-AIDS prevention or treatment. Another study showed that the holistic person-based approach is essential to achieve a change in behaviour . On the other hand, the WHO Report 2008  underlines the impact of person-based care in health improvement, quality of life, user trust and treatment adherence.
In this relationship model, the professional takes into account the values and perspective of the user, and therefore incorporates them in the decision-making process . The applicability of specific programmes requires a community approach to adjust them to match the needs of the target population [39, 42].
The results confirm that the AIDS-NGOs perform most of their activities in relation to target groups at risk of social exclusion or because they are socially vulnerable. Moreover, it reduces social inequalities due to socioeconomic status, gender and social orientation. It also confirms the role of NGO-AIDS as a bridge acting as a "communications space" between health services and the population) and also with other services (legal, social, employment, etc.).
Evaluation is a continuous process that facilitates the identification of areas for improvement. It should also contribute to the recognition of tasks that have been satisfactorily carried out by professionals. The professionals belonging to the participating AIDS-NGOs of this study consider evaluation to be an activity which is relevant to them and share the evaluation needs. Therefore, there is concordance between the necessities detected by the HIV-AIDS program (Public Health Department) and those identified through a revision of international publications [1, 43–45].
In order to evaluate the objectives and activities of AIDS-NGOs, it is essential to use good indicators bringing together characteristics such as: acceptability, feasibility, reliability, sensitivity to change, validity and meaningful and possible communicability [19, 46].
The latest update of the "Compendium of Evidence-Based HIV Prevention Interventions", elaborated by the CDC, as well as the "Community guide" website, both show a broad range of evidence-based preventive activities. However, these activities are included in programs aimed at very specific groups (female condom skills training; many men, many voices; personalized cognitive risk-reduction counseling), while our study presents a broader evaluation and suggests multiple indicators to evaluate concrete activities, which can be applicable to diverse intervention programs and to different groups. Thus, this proposal of indicators can help provide new evidence about the effectiveness of specific activities according to the level of prevention and intervention .
Through the development of the first phase of this study, two more target groups of population were identified: inmates and immigrants. These two target populations had specific characteristics deserving of special consideration and their own preventive programs on the part of the AIDS-NGOs, and they were added as target populations for purposes of this study. Specific objectives and activities addressed to these groups were identified, and specific evaluation indicators proposed through the seven final Haddon matrices, available via the following link (http://www.gencat.cat/salut/depsalut/html/ca/dir2068/informefinal_actongsida2009.pdf) .
Our proposal agrees with some of the quality measures of HIV care proposed by the CDC and UNAIDS [1, 45, 47]. What is new and constitutes a strong point in our study is that we have added other quality indicators that measure user satisfaction (an essential aspect of the evaluation of service quality) and more psychosocial factors (legal advice) that can have an impact on the fight against inequalities and stigmatization [6, 48].
Another important aspect to note is that, by using the Haddon matrix, many preventive interventions and evaluation indicators have been identified and defined on the individual, external and social levels, while less have been directed towards the individual's immediate environment. These findings are consistent with those observed by Peiró et al . A possible interpretation of this phenomenon is the concern of AIDS-NGOs professionals about data confidentiality; in order to act in the immediate environment of the affected person, he or she must agree to share the problem with individuals in his or her intimate environment.
The consensus phase among AIDS-NGOs professionals and the research team to develop the indicators has meant the cooperation and participation of the professionals in this study and constitutes a key element in facilitating the acceptance and implementation of the improvements proposed in the evaluation. It is also worth mentioning that professionals belonging to the AIDS-NGOs that participated in the consensus phase make up a multidisciplinary team much like the AIDS-NGOs. This gives the proposed indicators wide variability and richness, encompassing all of the activities undertaken by the AIDS-NGOs [34, 35, 45].
The main limitation of this proposal of indicators is that it has not been applied. It would be important to share and disseminate it among professionals of the various programs and organizations that work with HIV in order to test its applicability, usefulness, validity and practical relevance.
A possible selection bias could also be argued, given that these indicators have been designed with the participation of professionals from AIDS-NGOs that work in Catalonia and are financed by this autonomous community's Department of Health. Nonetheless, as mentioned above, the enormous variability in the number of activities and proposed evaluation indicators minimizes this bias and raises their potential usefulness for AIDS-NGOs located in other geographical areas. Furthermore, one limitation of the study was not to have studied the views of other health providers that work with HIV-AIDS prevention or control-programmes (Primary Care, Public Health, Reproductive and Sexual Health, Faith-Based Organizations and Community Based Organizations). Although their contributions would have been of great interest, for reasons of practicality and insufficient resources this aspect could not be included.
We also want to mention that indicators for global results such as HIV mortality, rate of hospitalizations and approach to other health problems (diabetes, hypertension...) have not been proposed. However, the improvements in these epidemiological indicators can be attributed to multiple factors (improvements in security and effectiveness of drug treatments, healthcare, etc.) and not only to the AIDS-NGOs' activities. Nonetheless, studies that directly involve classic systems of epidemiological surveillance for HIV-AIDS would be of interest.
For the selection of indicators in future evaluations of preventive activities in AIDS-NGOs, the elements that make up the third dimension of the Haddon matrix should be taken into consideration: effectiveness, cost, freedom, equity, stigmatization, preferences of the affected community or individual and viability. These factors can be considered from different points of view and can have greater or lesser weight when selecting indicators, based on the general objectives of future health policies and on HIV-AIDS [12, 15, 16].
It is also important that the strategic plan of future HIV-AIDS policies emphasize activities that address the population that makes up the affected or at-risk person's immediate environment . Although activities directed towards the social environment can have an indirect impact and their execution can be more difficult in some cases, it is necessary to emphasize the key role of the AIDS-NGOs in the development of these activities and the important social impact they can have.