Evidence-Based Practice in Social Work and Nursing

PART 1 – Searching as a craft: terms and context

Where searching sits in the cycle

In Module 2 you built a PICO question from a vague practice question. Now comes the step at which many students notice for the first time how much craft there is in academic work: finding. Evidence-based practice in nursing is described as a sequence of steps that runs from shared decision-making and problem formulation through searching and critical appraisal to a change in practice and evaluation. Searching therefore sits between the question and the critical appraisal: it supplies the material that you read critically in Module 5.

Keep the framework from Module 1 in mind. According to the Institute for Quality and Efficiency in Health Care (IQWiG, 2025), an evidence-based decision consists not only of evidence, but also of a person’s condition and circumstances and of their preferences. Searching therefore answers only part of the decision. It can show what research reports on a question; whether it fits this person, their life situation and their wishes is something you clarify later, in the team and with those involved (more on this in Module 7).

Key terms, clearly defined

Technical terms quickly seem arbitrary in term papers if they are not defined. So here are the basic terms of this module first. By literature search we mean here the planned and traceable search for publications on a clearly formulated question. A literature database is a searchable directory of such publications. A citation is the bibliographic reference to a text, an abstract is its short summary. According to the National Library of Medicine (National Library of Medicine [NLM], n.d.), PubMed, for example, contains more than 40 million citations and abstracts of biomedical literature (retrieved October 2026).

The search line (often also called a search string) is the combination of search terms and operators that you enter into a database. The search log records what you searched, where, when and with which line. The results list is what the database then returns. You also need a distinction that already played a role in Module 3: primary studies report on their own data collection, syntheses summarise several studies. Syntheses include systematic reviews, guidelines and expert standards. According to Cochrane Germany (n.d.), Cochrane, for example, is an international network that has been producing systematic reviews for more than 30 years. This module clarifies where you find such syntheses.

Searching broadly or narrowly: a trade-off

Every search sits between two wishes. A broad search finds a lot that is relevant, but also a lot that is not. A narrow search gives you a manageable list, but risks gaps. There is no setting that is always right. For a short presentation you may search more narrowly and look at overview articles first. If, on the other hand, you want to justify a practice question in the team, a broader search across several sources is worthwhile. What matters is that you know your choice and can justify it.

Figure 1 summarises the search routes of this module as a mind map: databases, evidence portals and syntheses tell you where to search; search strategy and documentation tell you how to search and how to make it traceable. The five branches also form the outline of the next parts.

PART 2 – Where to search: databases, evidence portals and syntheses

There is no single search engine for evidence-based practice. A different place is worthwhile depending on the question. Figure 2 sorts three types of access points; Table 1 compares them by content, access and suitability.

PubMed: a records database for biomedical literature

PubMed is run by the National Center for Biotechnology Information at the U.S. National Library of Medicine (National Library of Medicine [NLM], n.d.). Its focus is biomedical literature. That suits questions from health and nursing; for purely social-pedagogical questions you should also draw on other sources. A second point is often overlooked: PubMed itself does not contain full texts, but often links to them. So first you find the record, and then you have to obtain the text. Many full texts are only accessible via the university library (Hochschulbibliothek); others are freely available because they were published as open access. Allow time for obtaining them when you write term papers.

Cochrane and Cochrane Germany: reviews with plain-language summaries

Cochrane is an international network that has been producing systematic reviews for more than 30 years. Cochrane Germany has existed since 1998; since 2017 it has been a non-profit foundation funded by the Federal Ministry of Health and the University Medical Centre Freiburg (Cochrane Deutschland, n.d.). What is particularly useful for you as a student is that the reviews have plain-language summaries. They make it easier to get started, but they do not replace reading the review itself if you want to support a statement in a piece of work. The Cochrane Handbook (Higgins et al., 2024) is also the methodological reference to which the PICO mnemonic from Module 2 relates. The DNQP, too, names the Cochrane Library as one of its search sources (DNQP, 2025b).

Campbell Collaboration: evidence for social policy questions

For social work, the Campbell Collaboration is an important counterpart to the sources with a medical orientation. According to Welch (2021), it sees itself as a member-supported open-access organisation that provides systematic reviews and evidence and gap maps for social policy decisions. An Evidence and Gap Map is an overview that shows where studies exist on a topic area and where there are gaps. This helps you with the question of whether it is worth searching for effectiveness studies at all. In 2021 a new Coordinating Group “Child Welfare” was set up; alongside it there is, among others, the group “Social Welfare”. For questions from child and youth welfare this is an obvious starting point.

Finding guidelines and expert standards as syntheses

Sometimes you do not need individual studies, but guidance for practice. Then it is worth looking at guidelines and expert standards. The AWMF rules and regulations for guidelines (AWMF-Regelwerk Leitlinien) are available in version 2.2 of 22 May 2025. There, the “S” of the classification stands for the degree of systematic approach in the development process (AWMF, 2025). An S1 recommendation for action arises from the informal consensus of an expert group. S2k requires a representative guideline group and structured consensus-building under neutral facilitation, S2e a systematic search, selection and appraisal of the evidence. An S3 guideline combines all elements of systematic development. You practise how to classify guidelines in more depth in Module 6; for now it is enough to know this: do not search only for the topic, but read the class and the date as well.

According to the methods paper, expert standards of the German Network for Quality Development in Nursing (Deutsches Netzwerk für Qualitätsentwicklung in der Pflege, DNQP) are evidence-based, single-discipline instruments that describe nursing’s contribution to key quality risks (DNQP, 2025b). So they are not studies, but syntheses with a defined development procedure. The content of individual guidelines and standards is not the subject of this module; you learn how to find and classify them.

PART 3 – From the PICO question to the search line: building a search strategy

The route at a glance

According to the Cochrane Handbook, PICO is a mnemonic for populations, interventions, comparisons and outcomes, used to define the question and the inclusion criteria (Higgins et al., 2024). You know it from Module 2. Figure 3 shows the path from this question to the log in six steps.

Deriving search terms from PICO

You turn each building block of your question into a list of search terms. People do not all use the same words: what you call a “Wohngruppe” (residential group) is called “stationäre Jugendhilfe” (residential youth welfare) in other texts, or “residential care” in English. For each building block, therefore, collect synonyms, spelling variants (singular and plural, British and American spelling) and English technical terms. The latter are important because the large databases are indexed mainly in English. A good source of terms is the titles and abstracts of the first relevant hits: there you see how authors actually name the topic.

Not every PICO building block belongs in the search line. You often start with population, setting and intervention. Comparison and outcomes are often named differently or not at all in titles and abstracts; if you put them into the line too early, you lose relevant studies. They help you instead when screening: you can use them to recognise whether a hit fits your question.

Boolean operators: AND, OR, NOT

Boolean operators link search terms logically. Many interfaces expect them in capital letters; the exact spelling differs from database to database. The rule for everyday use: OR within a building block, AND between the building blocks, NOT only with great care. Table 2 compares the three operators.

Why is NOT risky? If you exclude a term, you also remove every hit in which it appears only in passing, for example in a list in the abstract. You do not notice what you lose as a result. For beginners the rule is: narrow the search with AND and filters instead, and sort out hits during screening.

Wildcards, phrases and filters

With a wildcard (truncation, often written as an asterisk) you search for all word forms with the same beginning: adolescent* finds adolescent, adolescents and adolescence. This saves typing, but it can also pick up unusable words if the word stem is too short. With quotation marks you mark a phrase, that is, words that are to be searched for together, for example "residential care". How wildcards and phrases work together differs slightly in each database. The help page tells you.

Filters limit the results list by properties such as period of publication, language or study type. They are practical, but never neutral: if you allow only the last few years, you give up older foundational work; if you allow only one language, you overlook studies from other countries. Note every filter in the log and justify it. Figure 4 shows how a complete search line results from synonym blocks: within the blocks OR widens, between the blocks AND narrows.

PART 4 – Working traceably: search log, screening and PRISMA flow diagram

The search log: your memory for the search

Without a log, in four weeks you will no longer know which line you entered where. More important still: nobody can trace your search. A log does not have to be elaborate, but it should record each search so that someone else could repeat it. Table 3 lists the fields that are enough to begin with.

Screening: from the results list to the included studies

The results list is only the raw material. When screening you proceed in stages. First you read the title and abstract and sort out what is clearly unsuitable. Then you obtain the full texts of the remaining candidates and appraise them more closely. What is left is included. It makes sense to derive inclusion and exclusion criteria from your PICO question in advance: Does the population fit? Is the intervention studied? Is there a comparison? Is a relevant outcome reported? For each exclusion at full-text stage, briefly note the reason. This costs little time and saves you later from the question of why a known study is missing.

PRISMA 2020: what the guideline does – and what it does not

PRISMA 2020 is a reporting guideline for systematic reviews. According to Page et al. (2021), it comprises a checklist of 27 items in seven sections and shows what should be reported transparently in a review. It is expressly not a guide to conducting a review and not an instrument for quality appraisal. This needs to be kept clearly apart: PRISMA says how a search is reported, not how good the studies found are. You learn how to appraise studies in Module 5. A core element is the flow diagram, which shows how many hits dropped out at which stage. Figure 5 shows our own, deliberately simplified version with the stages hits, screening and inclusion. The placeholders stand for the numbers that you determine in your own search.

PRISMA 2020 and the simplification shown here are not a compulsory programme for term papers or bachelor’s theses. They are reporting guidelines for publications. Even so, it is worth adopting their logic: if you know how many hits your own search produced, what was sorted out and why, you can justify your work better. Whether and to what extent you present the search path in your work depends on the requirements of your university and of the person supervising your work.

CLEANING Without duplicates SCREENING Title and abstract read FULL-TEXT REVIEW Full texts read INCLUSION Included studies Removed Duplicate records Excluded Does not fit the question Excluded Reasons noted

Figure 5: Simplified flow diagram “Hits – screening – inclusion”. Source: own illustration (AI-assisted), based on Page et al. (2021).

How a professional organisation searches: the DNQP example

The DNQP methods paper shows what this can look like in practice. It describes a hierarchical search: first systematic reviews, then primary studies, including in MEDLINE, CINAHL and the Cochrane Library. Documentation is by a PRISMA flow diagram; appraisal uses instruments such as AMSTAR-2, RoB 2 and AGREE II, and the trustworthiness of the evidence is presented using GRADE (DNQP, 2025b). You do not need to know these instruments now. For searching, the logic is what matters: first look for syntheses, then go into the individual studies, and document every step. That is the same order that you used in the worked example. In Module 3 you also saw that study designs and syntheses have different tasks; searching connects the two.

PART 5 – Limits, tensions and consequences

Good technique does not make a good search. Finally, it is about what even the most careful search cannot do.

No hits does not mean: no answer

If you find hardly any studies on a social-pedagogical question, this is first of all a finding about the data situation, not about the effect of a method. It can mean that there are none, that you searched the wrong source, that your terms do not match the literature, or that the question is studied in a form you did not expect. Biomedical databases such as PubMed cover social work questions only in part; that is why sources such as the Campbell Collaboration, with its evidence and gap maps, are so valuable. Missing studies imply neither that a measure works nor that it does not work. A “nothing found” belongs in the log together with the search path, not in a claim about effectiveness.

Whoever searches also decides: bias in the search process

A search is never completely objective. You decide which databases you use, which terms you choose, which filters you set and which hits you sort out. Someone who uses only English terms sees different texts from someone who also searches in German. Someone who searches only for randomised studies will hardly find qualitative studies on experience and meaning, even though in Module 3 these were classed as valuable in their own right. The log does not remove these distortions, but it makes them visible and so open to discussion. This fits with the following point: a search for a term paper is not a systematic review. Systematic reviews are produced with a fixed procedure and, as a rule, in a team; PRISMA describes how they are reported, not how to carry them out (Page et al., 2021). This module does not make you a review author, but a careful user of search routes. Module 8 shows you how a literature report differs from a systematic review.

Access, language and fairness

Anyone with access to a university library gets full texts more easily than professionals without that access. This is a fairness problem of science that reaches into practice: teams in small organisations sometimes have to manage without access to paid databases. Open-access offerings such as those of the Campbell Collaboration (Welch, 2021) and freely available methods papers such as the DNQP’s (2025b) ease this, but do not solve it. Language is a further issue: a large part of research is in English, while practice is in German. Translation errors and differences in context are a source of misunderstandings that you should watch for when reading.

Searching in the debate on evidence-based social work

The question of what counts as robust knowledge is contested in German-language social work; you know the supportive, critical and mediating positions of the professional debate from Module 1. For searching this means: what you search for reflects what you consider worth knowing. Anyone who searches only for effectiveness studies takes a position, even if this remains unspoken. A reflective search names this decision and asks whether the question might also need other types of study.

Consequences for study and practice

For your studies this means: describe your search so exactly that others can repeat it, and state its limits openly. For practice: do not search alone, but in the team and with a view to the people concerned. A good search supplies material for a decision. You make the decision itself with your professional knowledge and the preferences of those involved, following the procedures of your organisation. This module is not legal advice and replaces neither a lecture nor a practical placement.

Self-reflection – 5 questions

Voluntary · not evaluated – your answers are not saved and nobody analyses them. Click an answer; the explanation helps you to judge your own understanding.