Evidence review

Elderberry for Cold

Sambucus nigra

The short answer

Early or mixed evidence

Possibly a little for cold symptoms started early, but the evidence is mixed and product-specific. Small Sambucol trials and a 2019 meta-analysis of 180 people reported shorter, milder upper respiratory symptoms, and one air-travel trial found shorter colds without fewer of them. The best-designed trial, in 87 people with laboratory-confirmed influenza, found no benefit. EMA accepts elderflower as a traditional cold remedy but could not establish a berry monograph. Use cooked ripe-fruit products only; never raw berries.

7 key studies reviewed

Updated November 16, 2026

Key points

  1. 01

    Early trials looked promising

    In small Sambucol trials, flu-like symptoms improved about two to four days sooner than with placebo. Samples were tiny and methods imperfect.

  2. 02

    Prevention is unproven

    In 312 air travelers, elderberry did not clearly prevent colds, though colds that occurred were shorter and milder. A 2021 review rated that finding very uncertain.

  3. 03

    Best flu trial found no benefit

    In 87 emergency patients with PCR-confirmed influenza, elderberry extract did not shorten illness versus placebo, contradicting earlier studies.

  4. 04

    Flower is not berry

    EMA accepts elderflower teas as a traditional early-cold remedy but could not establish a berry monograph. Sinupret uses elder flower, not berry.

  5. 05

    Cook ripe fruit; check first if immunosuppressed

    Raw or unripe berries, leaves, and stems can release cyanide and cause vomiting. EMA cautions about immunosuppressants; avoid in pregnancy and under 18.

Elderberry syrups and gummies are a familiar first reach at the start of a cold or flu-like illness, but the clinical picture is genuinely mixed. Small trials of Sambucol, a standardized black elderberry (Sambucus nigra) syrup, reported influenza-like symptoms clearing about two to four days sooner than with placebo. A 2016 trial in air travelers found shorter, milder colds without clearly fewer of them, and a 2019 meta-analysis of 180 people reported a large benefit for upper respiratory symptoms. Then a carefully designed 2020 trial in 87 emergency patients with laboratory-confirmed influenza found no benefit at all. The National Center for Complementary and Integrative Health (NCCIH) says preliminary research suggests elderberry may relieve cold and flu symptoms, while a 2021 systematic review rated the evidence very uncertain and concluded that elderberry may not prevent colds.

Two plant parts matter, and they are not interchangeable. The purple-black berries are what most commercial “elderberry” cold products use and what the trials above tested. The creamy flowers, sold as elderflower, are a separate herbal drug: the European Medicines Agency (EMA) accepts elderflower teas and alcoholic preparations as traditional herbal medicines for early common-cold symptoms, based on long use rather than trials, but in 2014 its herbal committee concluded that a Community herbal monograph on the fruit (berry) could not be established. Sinupret, a German five-herb sinus medicine, contains elder flower, not berry. Cooked ripe berries as food are a different matter from concentrated supplements, and raw or unripe berries, leaves, and stems can cause cyanide-related poisoning.

This page walks through the key berry trials and who funded them, what pooled reviews concluded, how berry products differ from elderflower and from sinus formulas that use the flower, how trial doses map onto real products, and who should avoid elderberry. Nothing here is medical advice. A cold usually improves in one to two weeks with rest and fluids; shortness of breath, a high fever that lasts, symptoms that worsen or last too long, or illness in an infant needs medical care rather than a syrup.

02 The trials

Key studies

The trials and reviews that carry the most weight for this question, with the product and dose each one tested. Results apply to the preparation studied.

  1. Zakay-Rones et al., 1995

    Placebo-controlled, double-blind trial during an influenza B outbreak; randomization method not clearly reported

    Who
    27 people aged 5–56 on an Israeli kibbutz during a 1993 influenza B/Panama outbreak (15 elderberry, 12 placebo)
    What and how much
    Sambucol standardized elderberry extract syrup; 1 tablespoon twice daily for children, four times daily for adults
    How long
    3 days of treatment with symptom recording over 6 days

    A significant improvement, including fever, was seen in 93.3% of the Sambucol group within 2 days, versus 91.7% of the placebo group within 6 days (p < 0.001). Nearly 90% of the Sambucol group felt completely cured within 2–3 days, versus at least 6 days with placebo. The 2021 systematic review counted only 27 participants analyzed and rated the trial at unclear risk of bias because randomization was not described.

  2. Zakay-Rones et al., 2004

    Randomized, double-blind, placebo-controlled; sponsored by Razei Bar, Jerusalem

    Who
    60 adults aged 18–54 in Norway with influenza-like symptoms for 48 hours or less (mean about 27 hours)
    What and how much
    Sambucol syrup containing 38% proprietary elderberry extract, 15 ml four times daily, versus matching placebo
    How long
    5 days of treatment

    Symptoms were relieved on average 4 days earlier with elderberry than with placebo, and use of rescue medication was significantly lower. The authors asked for a larger confirmatory study. The 2021 review pooled this trial with the 1995 study and found recovery nearly three days faster, with very low certainty and inconsistent results.

  3. Tiralongo et al., 2016

    Randomized, double-blind, placebo-controlled; sponsored by Iprona AG, Italy

    Who
    312 economy-class air travelers from Australia on long-haul flights (mean age 51)
    What and how much
    Membrane-filtered Sambucus nigra extract capsules (300 mg, standardized to 15% anthocyanins and 22% polyphenols): 600 mg/day for 10 days before travel, then 900 mg/day until 4–5 days after arrival
    How long
    15–16 days depending on travel

    Cold episodes occurred in 12 elderberry and 17 placebo participants (not statistically significant, p = 0.4). Total cold-episode days were 57 versus 117 (p = 0.02), and symptom scores were lower with elderberry (247 versus 583, p = 0.05). Among people who caught a cold, mean duration was about 4.75 versus 6.88 days. The 2021 review rated prevention as very low certainty and concluded elderberry may not reduce the risk of catching a cold.

  4. Hawkins et al., 2019

    Meta-analysis of 4 randomized, controlled trials of elderberry for upper respiratory symptoms

    Who
    180 participants (89 elderberry, 91 placebo) across trials published 1995–2016
    What and how much
    Commercially prepared black elderberry (Sambucus nigra) products; doses not pooled in detail
    How long
    Varied by included trial

    Elderberry substantially reduced upper respiratory symptoms versus placebo, with a large standardized mean difference of 1.717 (95% CI 0.840–2.593). Heterogeneity was high (I² = 83%). Effects appeared stronger for influenza than for the common cold. The later 2021 systematic review cautioned that Downs and Black checklists understate risk of bias and that GRADE certainty is low or very low; this meta-analysis also preceded the negative 2020 emergency-department influenza trial.

  5. Macknin et al., 2020

    FDA-approved, investigator-initiated, double-blind, randomized, placebo-controlled; NCT03410862; powered >0.90 to detect a 2-day benefit

    Who
    87 patients aged 5 and older at three Midwestern emergency departments (January 2018–April 2019) with PCR-confirmed influenza, symptoms under 48 hours, and at least two moderate symptoms
    What and how much
    Elderberry extract 15 ml (5.7 g) twice daily for ages 5–12 or four times daily for older patients, versus placebo, for 5 days; oseltamivir allowed at patient choice
    How long
    5 days of treatment; outcomes followed until resolution or 21 days

    Time to all symptoms none or mild for 21.5 hours averaged 5.3 days with elderberry and 4.9 with placebo (p = 0.57). Complete resolution averaged 8.6 versus 8.7 days (p = 0.87). A post hoc analysis suggested elderberry alone without oseltamivir was about 2 days worse than placebo alone. The authors concluded there was no evidence of benefit and that results contradicted earlier studies.

  6. Harnett et al., 2020

    Rapid review of clinical studies of mono-herbal Sambucus nigra berry preparations

    Who
    Five clinical studies involving 936 adults with acute respiratory viral infections
    What and how much
    Commercial mono-herbal elderberry berry extracts, syrups, capsules, or lozenges (not combinations)
    How long
    Short courses up to 16 days across included studies

    Taken within 48 hours of symptom onset, berry preparations may reduce duration and severity of common cold and influenza symptoms. Within 2–4 days, most adult participants had about a 50% reduction in symptoms. Effects on cough and rescue-medicine use were inconsistent. Adverse events were rare and not serious. No reliable evidence for pregnancy, breastfeeding, or COVID-19.

  7. Wieland et al., 2021

    Systematic review with Cochrane risk-of-bias assessment and GRADE certainty ratings

    Who
    Five randomized trials of elderberry for prevention or treatment of viral respiratory illness (1,187 records screened)
    What and how much
    Elderberry alone or, in one trial, an echinacea–elderberry hot drink versus oseltamivir
    How long
    Literature search through early pandemic period; included trials lasted days to about 2 weeks

    Elderberry may not reduce the risk of developing a cold. It may reduce cold duration and severity and influenza duration, but evidence is uncertain (low or very low certainty). No studies linked elderberry to clinical inflammatory outcomes or a cytokine storm. Three studies measured cytokines after ingestion; none supported immune overstimulation in people. The authors judged elderberry a possibly safe option for viral respiratory illness, with uncertain benefits and harms.

03 Evidence

What the research shows

Early or mixed evidence

Our rating for Elderberry for Cold, based on the studies below.

The early positive story rests on two small Sambucol trials led by Zakay-Rones. In 1995, during an influenza B/Panama outbreak on an Israeli kibbutz, 27 people aged 5–56 took Sambucol or placebo. A significant improvement, including fever, was reported in 93.3% of the elderberry group within two days, against 91.7% of the placebo group within six days, and nearly 90% of the elderberry group felt cured within two to three days. The product also inhibited several influenza A and B strains in cell culture. In 2004, 60 Norwegian adults aged 18–54 with flu-like symptoms for 48 hours or less took 15 ml of Sambucol syrup (38% extract) or placebo four times a day for five days; symptoms eased about four days earlier and rescue medicines were used less. Razei Bar of Jerusalem sponsored that trial, and the authors themselves called for a larger confirmatory study. A 2009 Chinese pilot of proprietary lozenges (175 mg extract four times daily for two days in 64 adults), summarized in later reviews and sponsored by HerbalScience Singapore, also reported faster relief of fever, aches, and congestion, but it was not indexed on PubMed and did not report overall illness duration.

Evidence for ordinary colds is thinner and mostly from one travel trial. Tiralongo and colleagues (Nutrients 2016) randomized 312 Australian economy-class passengers on long-haul flights to a membrane-filtered Sambucus nigra extract (600 mg a day for ten days before travel, then 900 mg a day until four or five days after arrival) or placebo. Slightly fewer people in the elderberry group caught a cold (12 versus 17), a difference that was not statistically significant. Among those who did, total cold-episode days were 57 versus 117, symptom scores were lower, and mean duration was about two days shorter (4.75 versus 6.88 days). Iprona AG of Italy, the extract’s maker, sponsored the study. Harnett and colleagues’ 2020 rapid review of five studies with 936 adults concluded that mono-herbal berry preparations started within 48 hours may shorten and soften cold and flu symptoms, with most participants’ symptoms falling by about 50% within two to four days, while cough and rescue-medicine findings were inconsistent.

Pooled analyses look more favorable until the best-controlled flu trial is included. Hawkins and colleagues (Complementary Therapies in Medicine 2019) meta-analyzed four randomized trials with 180 participants and reported a large standardized mean difference of 1.717 (95% CI 0.840–2.593) for upper respiratory symptoms, with high heterogeneity and a stronger signal for influenza than for cold. That review did not include Macknin and colleagues’ 2020 emergency-department trial (Journal of General Internal Medicine). In that FDA-approved, investigator-initiated study, 87 patients aged 5 and older with polymerase chain reaction (PCR)–confirmed influenza and symptoms under 48 hours took elderberry extract 15 ml (5.7 g) twice a day (ages 5–12) or four times a day (older patients), or placebo, for five days, with oseltamivir (Tamiflu) allowed by choice. Time to mild symptoms averaged 5.3 days with elderberry and 4.9 with placebo; complete resolution took about 8.6 versus 8.7 days. A post hoc analysis suggested elderberry alone without oseltamivir was about two days worse than placebo alone. The trial was powered to detect a two-day benefit and found none.

Independent reviewers place the benefit, if any, as uncertain and product-specific. Wieland and colleagues (BMC Complementary Medicine and Therapies 2021) screened 1,187 records, included five randomized trials, and used GRADE: elderberry may not prevent colds; it may reduce cold duration and severity and influenza duration, all at low or very low certainty. They found no studies linking elderberry to clinical inflammatory harm or a “cytokine storm,” the COVID-era worry that grew out of laboratory work in which Sambucol products raised inflammatory messengers in white blood cells. Three human studies of cytokines after elderberry did not show harmful overstimulation. Compared with oseltamivir, an echinacea–elderberry hot drink performed roughly similarly by day five in one trial, but that tested a combination, not berry alone. EMA’s 2014 assessment of the fruit judged published clinical studies insufficient for well-established use and found traditional-use data too incomplete for a berry monograph, while its 2018 elderflower monograph grants only traditional use for early cold symptoms. Open questions remain: whether any benefit survives in larger independent trials of laboratory-confirmed influenza, how much product quality and standardization matter, and whether children’s syrups and gummies work at all.

04 In practice

How it was taken in studies

Trial doses apply only to the products tested. The Sambucol influenza trials used 15 ml of syrup four times a day for five days, started within 48 hours of symptoms; the 2004 syrup was 38% extract. The 2020 emergency trial used 15 ml (5.7 g) of extract four times a day for people over 12 and twice a day for ages 5–12, also for five days, without benefit. The air-travel trial used 300 mg capsules standardized to 15% anthocyanins and 22% polyphenols, two a day before travel and three a day during and after. Follow the label of a named product rather than converting between brands. For elderflower tea, EMA’s traditional posology is 2–5 g of dried flowers in 150 ml of boiling water three times a day (or related liquid extract and tincture doses) in adults and adolescents over 12; that is not a berry dose.

Elderberry sits alongside ordinary cold care, not instead of it. The UK National Health Service (NHS) advises rest, plenty of fluids, healthy food, and pharmacy advice on pain and fever relief; honey in a warm drink is an option only after age 1. Antibiotics do not help viral colds. Elderberry is not a flu vaccine and is not a substitute for prescription antivirals when a clinician recommends them for people at higher risk of flu complications. Do not stack several “immune” products that each contain elderberry, echinacea, zinc, or vitamin C, which makes dosing and side effects impossible to judge. If symptoms are no better after about a week on a product, or they worsen, stop relying on the syrup and seek care.

Choose cooked ripe-fruit products that name Sambucus nigra berry or fruit and state the extract strength. A 2023 Botanical Adulterants Prevention Program bulletin found that roughly one in six analyzed elderberry extract samples failed identity tests, most often through partial or complete substitution with cheaper black rice extract, which can cost up to about 30 times less than genuine elderberry extract; other anthocyanin-rich extracts and food dyes also appeared. Black rice is not poisonous, but it is not the studied berry. Check syrups and gummies for sugar if you manage diabetes, and store gummies like medicines, out of children’s reach. Homemade juices from raw berries crushed with stems are the riskiest preparations.

05 Caution

Before you try elderberry

Cooked, ripe elderberry products have been well tolerated in short clinical trials. Harnett’s review found adverse events rare and none serious, and when reported they were often at least as common with placebo. MSKCC lists infrequent gastrointestinal upset and immediate (type 1) allergy. EMA’s elderflower monograph lists no known undesirable effects. Trials have mostly lasted days to a few weeks, so months of daily high-dose extract are the least studied pattern. Stop the product and get urgent help for swelling of the face or throat, widespread hives, or difficulty breathing.

Raw plant material is a different hazard. NCCIH warns that raw or unripe elderberries and other parts of the plant, such as leaves and stems, contain cyanide-producing substances that can cause nausea, vomiting, and severe diarrhea, and that cooking eliminates the toxin. In August 1983, after a California religious group drank juice pressed from wild elderberries crushed with leaves and branches, 11 of 25 people became ill within 15 minutes; eight were flown by helicopter to hospital with vomiting, cramps, and weakness, and one who had drunk five glasses was hospitalized overnight. All recovered; serum cyanide levels were normal, a reminder that other plant toxins may also matter. EMA states that plant parts other than flowers and ripe berries are poisonous, that leaves and stems should not be crushed into preparations, and that heat treatment of ripe fruit reduces the small amount of hydrogen cyanide seeds can release.

EMA does not recommend berry preparations in pregnancy or breastfeeding, or under age 18, because safety and efficacy data are lacking; elderflower medicines are not recommended under 12 or in pregnancy or lactation for the same reason. NCCIH says little is known about medicinal elderberry in pregnancy and breastfeeding, and MSKCC advises avoiding it then. EMA also notes that elderberry might, in theory, interfere with immunosuppressant drugs and corticosteroids because of possible immunostimulating activity seen in laboratory cytokine studies. Clinical evidence of such harm is lacking, and Wieland’s review found no evidence of immune overstimulation in people, but MSKCC lists a case of autoimmune hepatitis in a woman with Hashimoto’s thyroiditis who took elderberry-containing supplements long term. People with autoimmune disease, organ transplants, or immunosuppressant prescriptions should ask their specialist before using elderberry.

Most colds settle at home, but NHS advice sets clear red flags. See a GP if you have a high temperature for more than 3 days, symptoms get worse, you feel short of breath or have chest pain, symptoms do not improve after 10 days, a cough lasts more than 3 weeks, or you have a long-term condition or a weakened immune system. For babies and young children, seek urgent help for difficulty breathing, blue or gray lips, unusual sleepiness or floppiness, a fever of 38°C (100.4°F) or higher under 3 months of age, or 39°C (102.2°F) or higher at 3–6 months. Elderberry should never delay care when these signs appear.

06 Questions

Frequently Asked Questions

Possibly a little, if you start early, but do not count on it. Small elderberry syrup trials and a 2016 air-travel study reported shorter or milder cold symptoms, and Hawkins’s 2019 meta-analysis of 180 people found a large average benefit. A 2021 systematic review rated the evidence very uncertain and said elderberry may not prevent colds. Our guide to herbs for colds, flu and immunity compares other options.

The early Sambucol trials suggested flu-like symptoms cleared two to four days sooner, but the best-designed study so far did not. In 87 emergency patients with PCR-confirmed influenza, Macknin and colleagues found no difference versus placebo in time to mild symptoms or full recovery. Elderberry is not a flu vaccine and is not a substitute for antivirals when a clinician recommends them.

No. They come from the same tree but are different medicines. EMA accepts elderflower tea as a traditional remedy for early cold symptoms; it could not establish a monograph for the berry. Most clinical trials tested berry extracts. Sinupret uses elder flower (with four other herbs) for sinus symptoms, not berry syrup.

Many commercial blends already combine them, which makes it hard to know which ingredient helped or caused a reaction. Echinacea has its own mixed evidence for colds and its own cautions in autoimmune disease. Kan Jang (andrographis with a little eleuthero) and pelargonium are other researched cold options with different safety profiles. Check every label so you do not double-dose, and ask a pharmacist if you take prescription medicines.

Only cooked ripe fruit is considered safe as food. NCCIH warns that raw or unripe berries, leaves, and stems contain cyanide-producing substances that can cause nausea, vomiting, and severe diarrhea; cooking removes the toxin. The 1983 California outbreak followed juice pressed from wild berries crushed with leaves and branches. Never make homemade “immune shots” from raw berries and stems.

EMA does not recommend berry preparations under 18 or in pregnancy or breastfeeding, and advises checking first if you take immunosuppressants or corticosteroids. MSKCC advises avoiding medicinal elderberry in pregnancy and lactation. Anyone allergic to elder flowers, berries, or pollen should skip it. People with autoimmune disease or transplants should discuss it with their specialist before use.

See a doctor for a high fever lasting more than 3 days, shortness of breath, chest pain, symptoms that worsen or last beyond about 10 days, a cough lasting more than 3 weeks, or cold symptoms with a long-term illness or weak immunity (NHS). Get urgent help for infants with high fever or breathing difficulty. A syrup is never a reason to wait when these red flags appear.

07 References

Sources

These references support the claims on this page. They are not an endorsement of any product. The elderberry monograph carries its own full source list.

  1. Inhibition of several strains of influenza virus in vitro and reduction of symptoms by an elderberry extract (Sambucus nigra L.) during an outbreak of influenza B Panama

    Journal of Alternative and Complementary Medicine (PubMed 9395631), 1995

    Zakay-Rones et al.: Sambucol inhibited influenza strains in cell culture; in a 1993 kibbutz outbreak, 93.3% improved within 2 days versus 6 days with placebo; nearly 90% cured within 2–3 days.

  2. Randomized study of the efficacy and safety of oral elderberry extract in the treatment of influenza A and B virus infections

    Journal of International Medical Research (PubMed 15080016), 2004

    Zakay-Rones et al.: 60 Norwegian adults; Sambucol 15 ml four times daily for 5 days; symptoms relieved about 4 days earlier than placebo; larger study requested.

  3. Elderberry Supplementation Reduces Cold Duration and Symptoms in Air-Travellers: A Randomized, Double-Blind Placebo-Controlled Clinical Trial

    Nutrients (PubMed 27023596), 2016

    Tiralongo et al.: 312 travelers; cold episodes 12 versus 17 (not significant); cold days 57 versus 117; mean duration about 4.75 versus 6.88 days; sponsored by Iprona AG.

  4. Black elderberry (Sambucus nigra) supplementation effectively treats upper respiratory symptoms: A meta-analysis of randomized, controlled clinical trials

    Complementary Therapies in Medicine (PubMed 30670267), 2019

    Hawkins et al.: 180 participants across 4 RCTs; SMD 1.717 (95% CI 0.840–2.593); high heterogeneity; stronger signal for influenza than cold; preceded Macknin 2020.

  5. Elderberry Extract Outpatient Influenza Treatment for Emergency Room Patients Ages 5 and Above: a Randomized, Double-Blind, Placebo-Controlled Trial

    Journal of General Internal Medicine (PubMed 32929634), 2020

    Macknin et al.: 87 PCR-confirmed influenza patients; no benefit on time to mild symptoms (5.3 versus 4.9 days) or resolution; post hoc, elderberry alone about 2 days worse than placebo alone; NCT03410862.

  6. The effects of Sambucus nigra berry on acute respiratory viral infections: A rapid review of clinical studies

    Advances in Integrative Medicine (PubMed 32864330), 2020

    Harnett et al.: five studies, 936 adults; mono-herbal berry within 48 hours may reduce duration and severity; ~50% symptom reduction in 2–4 days; no reliable evidence in pregnancy or for COVID-19.

  7. Elderberry for prevention and treatment of viral respiratory illnesses: a systematic review

    BMC Complementary Medicine and Therapies (PubMed 33827515), 2021

    Wieland et al.: five RCTs; may not prevent colds; may shorten colds and influenza at low/very low certainty; no evidence of clinical immune overstimulation or cytokine storm.

  8. European Union herbal monograph on Sambucus nigra L., flos

    European Medicines Agency (EMA/HMPC/611512/2016), 2018

    Traditional use of elderflower for early common-cold symptoms; infusion/decoction and liquid extract/tincture doses; not under 12; not recommended in pregnancy or lactation; see a doctor after one week or with dyspnoea, fever, or purulent sputum.

  9. Assessment report on Sambucus nigra L., fructus

    European Medicines Agency (EMA/HMPC/44208/2012), 2014

    Clinical trials insufficient for well-established use; traditional-use data incomplete; Community herbal monograph on the fruit cannot be established; heat removes HCN; not under 18 or in pregnancy/lactation; possible interference with immunosuppressants.

  10. Elderberry

    National Center for Complementary and Integrative Health (NIH), 2024

    Preliminary research may support symptom relief for colds and flu; no good evidence for COVID-19; raw/unripe berries, leaves, and stems contain cyanide-producing toxins removed by cooking; little known in pregnancy and breastfeeding.

  11. Elderberry

    Memorial Sloan Kettering Cancer Center, About Herbs, 2025

    European elder most studied; avoid in pregnancy and lactation; raw plant cyanide warning; autoimmune hepatitis and pancreatitis case reports; theoretical diabetes-drug, diuretic, and laxative interactions.

  12. Poisoning from Elderberry Juice — California

    Centers for Disease Control and Prevention, MMWR 33(13), 1984

    August 1983: juice from wild berries crushed with leaves and branches; 11 of 25 ill, eight flown to hospital; serum cyanide normal; leaves and stems should not be crushed into juice.

  13. Adulteration of European Elder (Sambucus nigra) Berries and Berry Extracts

    ABC-AHP-NCNPR Botanical Adulterants Prevention Program (Gafner and Brinckmann), 2023

    About 17% of analyzed samples failed identity tests; black rice extract the most common adulterant, up to ~30 times cheaper; other anthocyanin extracts and food dyes reported.

  14. Common cold

    NHS, 2026

    Usually improves in 1–2 weeks; see a GP for fever over 3 days, worsening symptoms, shortness of breath or chest pain, no improvement after 10 days, cough over 3 weeks, or colds with long-term illness or weak immunity.

All evidence reviews →